Daily Papers

We propose T2I-ReasonBench, a benchmark evaluating reasoning capabilities of text-to-image (T2I) models. It consists of four dimensions: Idiom Interpretation, Textual Image Design, Entity-Reasoning and Scientific-Reasoning. We propose a two-stage evaluation protocol to assess the reasoning accuracy and image quality. We benchmark various T2I generation models, and provide comprehensive analysis on their performances.

We introduce ESGenius, a comprehensive benchmark for evaluating and enhancing the proficiency of Large Language Models (LLMs) in Environmental, Social and Governance (ESG) and sustainability-focused question answering. ESGenius comprises two key components: (i) ESGenius-QA, a collection of 1 136 multiple-choice questions generated by LLMs and rigorously validated by domain experts, covering a broad range of ESG pillars and sustainability topics. Each question is systematically linked to its corresponding source text, enabling transparent evaluation and supporting retrieval-augmented generation (RAG) methods; and (ii) ESGenius-Corpus, a meticulously curated repository of 231 foundational frameworks, standards, reports and recommendation documents from seven authoritative sources. Moreover, to fully assess the capabilities and adaptation potential of the model, we implement a rigorous two-stage evaluation protocol -- Zero-Shot and RAG. Extensive experiments across 50 LLMs (ranging from 0.5 B to 671 B parameters) demonstrate that state-of-the-art models achieve only moderate performance in zero-shot settings, with accuracies typically around 55--70\%, highlighting ESGenius's challenging nature for LLMs in interdisciplinary contexts. However, models employing RAG show significant performance improvements, particularly for smaller models. For example, "DeepSeek-R1-Distill-Qwen-14B" improves from 63.82\% (zero-shot) to 80.46\% with RAG. These results underscore the necessity of grounding responses in authoritative sources for enhanced ESG understanding. To the best of our knowledge, ESGenius is the first benchmark curated for LLMs and the relevant enhancement technologies that focuses on ESG and sustainability topics.

In panoptic scene graph generation (PSGG), models retrieve interactions between objects in an image which are grounded by panoptic segmentation masks. Previous evaluations on panoptic scene graphs have been subject to an erroneous evaluation protocol where multiple masks for the same object can lead to multiple relation distributions per mask-mask pair. This can be exploited to increase the final score. We correct this flaw and provide a fair ranking over a wide range of existing PSGG models. The observed scores for existing methods increase by up to 7.4 mR@50 for all two-stage methods, while dropping by up to 19.3 mR@50 for all one-stage methods, highlighting the importance of a correct evaluation. Contrary to recent publications, we show that existing two-stage methods are competitive to one-stage methods. Building on this, we introduce the Decoupled SceneFormer (DSFormer), a novel two-stage model that outperforms all existing scene graph models by a large margin of +11 mR@50 and +10 mNgR@50 on the corrected evaluation, thus setting a new SOTA. As a core design principle, DSFormer encodes subject and object masks directly into feature space.

Recent deployments of large language models (LLMs) as autonomous trading agents raise questions about whether financial decision-making competence generalizes beyond specific market patterns and how it should be trained and evaluated in noisy markets lacking ground truth. We propose a structured framework for training and evaluating such models. Central to our approach is a curated, multiple-choice question (MCQ) dataset derived from classic textbooks and historical markets, verified by an AI committee, enriched with structured reasoning traces, and augmented to reduce shortcut learning. To evaluate whether performance on isolated MCQs generalizes to real-world trading, we introduce a two-stage protocol combining test-set evaluation with an MCQ-based chronological trading simulation. Extensive evaluations across market regimes provide statistically robust evidence that open models trained with our framework exhibit competitive, risk-aware behavior over time, outperform open-source baselines, and approach frontier-model performance at smaller scale. We release the dataset and evaluation framework to support further research.

Vision-Language Models (VLMs) show promise for autonomous driving, yet their struggle with hallucinations, inefficient reasoning, and limited real-world validation hinders accurate perception and robust step-by-step reasoning. To overcome this, we introduce AgentThink, a pioneering unified framework that integrates Chain-of-Thought (CoT) reasoning with dynamic, agent-style tool invocation for autonomous driving tasks. AgentThink's core innovations include: (i) Structured Data Generation, which establishes an autonomous driving tool library to automatically construct structured, self-verified reasoning data explicitly incorporating tool usage for diverse driving scenarios; (ii) A Two-stage Training Pipeline, employing Supervised Fine-Tuning (SFT) with Group Relative Policy Optimization (GRPO) to equip VLMs with the capability for autonomous tool invocation; and (iii) Agent-style Tool-Usage Evaluation, introducing a novel multi-tool assessment protocol to rigorously evaluate the model's tool invocation and utilization. Experiments on the DriveLMM-o1 benchmark demonstrate that AgentThink significantly boosts overall reasoning scores by 53.91% and enhances answer accuracy by 33.54%, while markedly improving reasoning quality and consistency. Furthermore, ablation studies and robust zero-shot/few-shot generalization experiments across various benchmarks underscore its powerful capabilities. These findings highlight a promising trajectory for developing trustworthy and tool-aware autonomous driving models. Code is available at https://github.com/curryqka/AgentThink.

Following recent advancements in computer-aided detection and diagnosis systems for colonoscopy, the automated reporting of colonoscopy procedures is set to further revolutionize clinical practice. A crucial yet underexplored aspect in the development of these systems is the creation of computer vision models capable of autonomously segmenting full-procedure colonoscopy videos into anatomical sections and procedural phases. In this work, we aim to create the first open-access dataset for this task and propose a state-of-the-art approach, benchmarked against competitive models. We annotated the publicly available REAL-Colon dataset, consisting of 2.7 million frames from 60 complete colonoscopy videos, with frame-level labels for anatomical locations and colonoscopy phases across nine categories. We then present ColonTCN, a learning-based architecture that employs custom temporal convolutional blocks designed to efficiently capture long temporal dependencies for the temporal segmentation of colonoscopy videos. We also propose a dual k-fold cross-validation evaluation protocol for this benchmark, which includes model assessment on unseen, multi-center data.ColonTCN achieves state-of-the-art performance in classification accuracy while maintaining a low parameter count when evaluated using the two proposed k-fold cross-validation settings, outperforming competitive models. We report ablation studies to provide insights into the challenges of this task and highlight the benefits of the custom temporal convolutional blocks, which enhance learning and improve model efficiency. We believe that the proposed open-access benchmark and the ColonTCN approach represent a significant advancement in the temporal segmentation of colonoscopy procedures, fostering further open-access research to address this clinical need.

As the development of Large Models (LMs) progresses rapidly, their safety is also a priority. In current Large Language Models (LLMs) and Multimodal Large Language Models (MLLMs) safety workflow, evaluation, diagnosis, and alignment are often handled by separate tools. Specifically, safety evaluation can only locate external behavioral risks but cannot figure out internal root causes. Meanwhile, safety diagnosis often drifts from concrete risk scenarios and remains at the explainable level. In this way, safety alignment lack dedicated explanations of changes in internal mechanisms, potentially degrading general capabilities. To systematically address these issues, we propose an open-source project, namely DeepSight, to practice a new safety evaluation-diagnosis integrated paradigm. DeepSight is low-cost, reproducible, efficient, and highly scalable large-scale model safety evaluation project consisting of a evaluation toolkit DeepSafe and a diagnosis toolkit DeepScan. By unifying task and data protocols, we build a connection between the two stages and transform safety evaluation from black-box to white-box insight. Besides, DeepSight is the first open source toolkit that support the frontier AI risk evaluation and joint safety evaluation and diagnosis.

Test-Time Adaptation (TTA) has recently emerged as a promising approach for tackling the robustness challenge under distribution shifts. However, the lack of consistent settings and systematic studies in prior literature hinders thorough assessments of existing methods. To address this issue, we present TTAB, a test-time adaptation benchmark that encompasses ten state-of-the-art algorithms, a diverse array of distribution shifts, and two evaluation protocols. Through extensive experiments, our benchmark reveals three common pitfalls in prior efforts. First, selecting appropriate hyper-parameters, especially for model selection, is exceedingly difficult due to online batch dependency. Second, the effectiveness of TTA varies greatly depending on the quality and properties of the model being adapted. Third, even under optimal algorithmic conditions, none of the existing methods are capable of addressing all common types of distribution shifts. Our findings underscore the need for future research in the field to conduct rigorous evaluations on a broader set of models and shifts, and to re-examine the assumptions behind the empirical success of TTA. Our code is available at https://github.com/lins-lab/ttab.

We review human evaluation practices in automated, speech-driven 3D gesture generation and find a lack of standardisation and frequent use of flawed experimental setups. This leads to a situation where it is impossible to know how different methods compare, or what the state of the art is. In order to address common shortcomings of evaluation design, and to standardise future user studies in gesture-generation works, we introduce a detailed human evaluation protocol for the widely-used BEAT2 motion-capture dataset. Using this protocol, we conduct large-scale crowdsourced evaluation to rank six recent gesture-generation models -- each trained by its original authors -- across two key evaluation dimensions: motion realism and speech-gesture alignment. Our results provide strong evidence that 1) newer models do not consistently outperform earlier approaches; 2) published claims of high motion realism or speech-gesture alignment may not hold up under rigorous evaluation; and 3) the field must adopt disentangled assessments of motion quality and multimodal alignment for accurate benchmarking in order to make progress. Finally, in order to drive standardisation and enable new evaluation research, we will release five hours of synthetic motion from the benchmarked models; over 750 rendered video stimuli from the user studies -- enabling new evaluations without model reimplementation required -- alongside our open-source rendering script, and the 16,000 pairwise human preference votes collected for our benchmark.

Recent text-to-video (T2V) technology advancements, as demonstrated by models such as Gen2, Pika, and Sora, have significantly broadened its applicability and popularity. Despite these strides, evaluating these models poses substantial challenges. Primarily, due to the limitations inherent in automatic metrics, manual evaluation is often considered a superior method for assessing T2V generation. However, existing manual evaluation protocols face reproducibility, reliability, and practicality issues. To address these challenges, this paper introduces the Text-to-Video Human Evaluation (T2VHE) protocol, a comprehensive and standardized protocol for T2V models. The T2VHE protocol includes well-defined metrics, thorough annotator training, and an effective dynamic evaluation module. Experimental results demonstrate that this protocol not only ensures high-quality annotations but can also reduce evaluation costs by nearly 50%. We will open-source the entire setup of the T2VHE protocol, including the complete protocol workflow, the dynamic evaluation component details, and the annotation interface code. This will help communities establish more sophisticated human assessment protocols.

LLM agents increasingly modify their own prompts, tools, middleware, resources, and execution harnesses at runtime. Such self-evolution can improve capability, but a successful mutation may leave persistent effects that cannot be safely reversed in states different from the one in which it was created. We introduce EvoUndo, a framework for representing, synthesizing, diagnosing, and independently verifying recoverability of model-generated self-modifications across counterfactual states. Across 600 unseen one-shot self-evolution tasks, we identify 197 capability-improving mutations that fail recoverability verification. Under the original recovery representation, conventional repair strategies recover 0/197 of these natural failures. Deterministic oracle analysis recovers 48/197 under the original recovery language L0, while the extended recovery calculus increases empirical oracle recovery to 191/197. A protocol-locked 2x2 grounding-by-expressivity intervention then separates two bottlenecks: exact state-address grounding increases successful recovery from 0/48 to 38/48 (79.2%) when the original language is sufficient, while extending the recovery language enables recovery on 142/143 (99.3%) failures in the oracle-defined S1 stratum. On the primary gpt-oss-120b backbone, adding exact-address diagnostics to the richer language reduces recovery to 133/143 (93.0%); a Qwen3.8-27B replication preserves the grounding and expressivity effects but not this negative interaction, indicating that the latter is model-dependent. These results indicate that reliable agent self-evolution requires co-designing verification, state grounding, witness semantics, and recovery-language expressivity rather than relying on iterative prompting alone.

LeWorldModel trains a latent world model with a prediction loss and a single anti-collapse regulariser, and reports approximately 87% of goals reached on TwoRoom, its simplest diagnostic environment. We reproduce that result by independent reimplementation on roughly $25 of rented compute, with all evaluation on one laptop CPU. We reach 94.0% at the repository's evaluation goal offset, against 84.0% for the authors' own released checkpoint measured under our protocol on identical episodes, and we reproduce the reported representation result directly (position probe Pearson r = 0.9988 against a reported 0.996). Reaching that point required four conventions that determine the outcome and appear in no released configuration file: dense action gathering across a frameskip block, a programmatically-set action-encoder width, ImageNet pixel normalisation, and action z-scoring. A reproducer following the released configurations alone obtains a model whose predictor cannot converge. The evaluation protocol is itself contested by the released material. The paper's appendix and the repository's configuration specify different goal offsets and step budgets; on the authors' own weights these yield 14.0% and 84.0%, and only the configuration's values reproduce the reported figure. On fifty identical episodes, changing nothing but how the goal is constructed moves that checkpoint from 84.0% to 8.0%. Two findings generalise. One-step prediction accuracy does not predict long-horizon planning success: across three checkpoints spanning a sevenfold range in prediction error, including the authors' own, it orders short-horizon success monotonically and fails to order long-horizon success at all. And a batch normalisation layer inflated our reported validation loss by up to a factor of 300, concealing a training loss that was flat throughout.

Comprehensive and constructive evaluation protocols play an important role in the development of sophisticated text-to-video (T2V) generation models. Existing evaluation protocols primarily focus on temporal consistency and content continuity, yet largely ignore the dynamics of video content. Dynamics are an essential dimension for measuring the visual vividness and the honesty of video content to text prompts. In this study, we propose an effective evaluation protocol, termed DEVIL, which centers on the dynamics dimension to evaluate T2V models. For this purpose, we establish a new benchmark comprising text prompts that fully reflect multiple dynamics grades, and define a set of dynamics scores corresponding to various temporal granularities to comprehensively evaluate the dynamics of each generated video. Based on the new benchmark and the dynamics scores, we assess T2V models with the design of three metrics: dynamics range, dynamics controllability, and dynamics-based quality. Experiments show that DEVIL achieves a Pearson correlation exceeding 90% with human ratings, demonstrating its potential to advance T2V generation models. Code is available at https://github.com/MingXiangL/DEVIL.

Benchmarks are necessary for healthcare evaluation, but are not sufficient for predicting deployment performance. Our position is that the evaluation--deployment gap arises not because of poorly designed benchmarks, but from implicit assumptions about how users interact with models that cannot be surfaced from benchmarks alone. To make this precise, we propose a classification of assumptions into two categories: task, which can be tested from conversation data alone, and outcome, which requires outcome data and behavioral studies for testing. Critically, outcome assumptions depend on human behavior, something that even well-designed benchmarks cannot directly observe. To demonstrate the operationality of this framework, we retrospectively analyze a healthcare RCT as a case study and find that the gap naturally separates into task and outcome gaps of roughly equal size. To address this, we make two contributions: first, we propose BenchmarkCards, an artifact that documents assumptions, and second, we propose staged evaluation, a procedure that systematically tests assumptions and evaluates performance.

AI research agents combine prior knowledge, public sources, and experimental feedback to produce useful results. The Discovery Certification Protocol (DCP) turns claims about these results into executable recovery and feedback tests. Gate 1 validates useful improvement on sealed evaluation. Gate 2 gives matched agents the registered starting information and observed Web content while withholding the target research history. Every valid method reaching the numerical target supplies a recovery witness and triggers the Core veto. DCP Core requires adequate controls, zero observed recoveries, and a finite-sample bound on recovery in one fresh registered episode. Optional Gate 3 measures the average effect of truthful feedback relative to a specified neutral policy from a shared checkpoint. DCP Evidence adds this effect after independent null calibration and a registered effect margin. Two controlled audits exercise the complete protocol in SQLite optimization and virtual catalyst control under different models. Each produced zero recoveries in 96 episodes, with an upper bound of 0.0468. Each paired study yielded 30 truthful recoveries and zero neutral recoveries, with passing 60-pair null studies. Additional cases exercise Core, recovered, and audit-incomplete decisions. A deterministic, LLM-free verifier reproduces the decisions from frozen evidence. DCP provides a common evidence language for useful outcomes, alternative routes, and feedback effects across AI research.

Video summarization is a technique to create a short skim of the original video while preserving the main stories/content. There exists a substantial interest in automatizing this process due to the rapid growth of the available material. The recent progress has been facilitated by public benchmark datasets, which enable easy and fair comparison of methods. Currently the established evaluation protocol is to compare the generated summary with respect to a set of reference summaries provided by the dataset. In this paper, we will provide in-depth assessment of this pipeline using two popular benchmark datasets. Surprisingly, we observe that randomly generated summaries achieve comparable or better performance to the state-of-the-art. In some cases, the random summaries outperform even the human generated summaries in leave-one-out experiments. Moreover, it turns out that the video segmentation, which is often considered as a fixed pre-processing method, has the most significant impact on the performance measure. Based on our observations, we propose alternative approaches for assessing the importance scores as well as an intuitive visualization of correlation between the estimated scoring and human annotations.

The automatic evaluation of instruction following typically involves using large language models (LLMs) to assess response quality. However, there is a lack of comprehensive evaluation of these LLM-based evaluators across two dimensions: the base LLMs and the evaluation protocols. Therefore, we present a thorough meta-evaluation of instruction following, including 25 base LLMs and 15 recently proposed evaluation protocols, on 4 human-annotated datasets, assessing the evaluation accuracy of the LLM-evaluators. Our evaluation allows us to identify the best-performing base LLMs and evaluation protocols with a high degree of robustness. Moreover, our large-scale evaluation reveals: (1) Base LLM performance ranking remains largely consistent across evaluation protocols, with less capable LLMs showing greater improvement from protocol enhancements; (2) Robust evaluation of evaluation protocols requires many base LLMs with varying capability levels, as protocol effectiveness can depend on the base LLM used; (3) Evaluation results on different datasets are not always consistent, so a rigorous evaluation requires multiple datasets with distinctive features. We release our meta-evaluation suite ReIFE, which provides the codebase and evaluation result collection for more than 500 LLM-evaluator configurations, to support future research in instruction-following evaluation.

Large Language Models (LLMs) are widely used as proxies for human labelers in both training (Reinforcement Learning from AI Feedback) and large-scale response evaluation (LLM-as-a-judge). Alignment and evaluation are critical components in the development of reliable LLMs, and the choice of feedback protocol plays a central role in both but remains understudied. In this work, we show that the choice of feedback protocol for evaluation (absolute scores versus relative preferences) can significantly affect evaluation reliability and induce systematic biases. In the context of LLM-as-a-judge evaluation, we show that pairwise protocols are more vulnerable to distracted evaluation. Generator models can exploit spurious attributes (or distractor features) favored by the LLM judge, resulting in inflated scores for lower-quality outputs. We find that absolute scoring is more robust to such manipulation, producing judgments that better reflect response quality and are less influenced by distractor features. Our results demonstrate that generator models can flip preferences by embedding distractor features, skewing LLM-as-a-judge comparisons and leading to inaccurate conclusions about model quality in benchmark evaluations. Pairwise preferences flip in about 35% of the cases, compared to only 9% for absolute scores. We offer recommendations for choosing feedback protocols based on dataset characteristics and evaluation objectives.

Red-team evaluations of AI models support some claims and not others, and the boundary between the two is calculable rather than merely a matter of judgment. We define the evidential ceiling of an evaluation as the largest factor by which one result can move belief under a fixed testing budget, derive it in closed form for the benchmark null result, and use it to locate that boundary exactly. We find that above a calculable harm rate, a benchmark of modest size certifies a category to a stated evidentiary standard, and a clean sheet is then the stronger of the two possible observations, outweighing a single reproduced failure. Below that rate, no passive benchmark of feasible size provides the specified evidence of safety under the fixed scoring rule and approximately independent trial structure. The crossing between the two regimes has a closed form. The bound is not specific to benchmarks: written in terms of a procedure's hypothesis conditioned elicitation rates, it covers adaptive and automated red teaming as well, and shows that discrimination between the hypotheses rather than attack success is what determines evidential worth. Auditing eight evaluation suites against the boundary, we find that current benchmarks are adequate for high-frequency harm categories and several orders of magnitude short for rare, catastrophic ones. Safety benchmarks are not uninformative. They are informative about a specific and computable set of propositions, and the discipline they need is to state which.

Benchmark contamination is diagnosed with n-gram overlap, likelihood-based membership inference, or canary strings, and each needs something usually unavailable: the training corpus, a well-chosen test statistic, or foresight at release. A recent alternative reads it off a linear probe on internal activations. We show the natural way to do this does not work, specify one that survives measurement, then find that the correction making it work carries more variance than the null it is tested against. The protocol reports a zero-sum contrast on the depth profile of probe accuracy, recentred on a level-matched placebo baseline, tested against a label-permutation null, with the reference set twice the size of the suspect set. Each choice replaces a simpler alternative we rejected on measurement. Reporting the level of excess separability rather than its shape makes the false positive rate track the size of the analyst's own control set, 0.03 to 0.99 under a true null. Contrasting against a flat depth profile rejects a true null 0.72 of the time when surface decodability rises with depth, and loses all power when it falls. On real transformers the protocol fails a test the simulations did not pose. The recentring subtracts an estimate, and the permutation null holds it fixed. Re-estimated across split seeds on four audits of contaminated checkpoints, its standard deviation is 1.30 to 1.56 times the null's own in every arm: what is subtracted to remove a bias is more variable than what it corrects. The one nominally significant result, p = 0.0075, becomes 0.0745 once that variance is propagated, and no verdict is issued. The simulations missed this because their surface key is the covariate driving item variation; on real text it is a proxy, and degrading key quality in simulation reproduces it. We add a companion measurement and a widened null. No arm shows contamination.

Systems that generate scientific research questions are evaluated today by expert scores, LLM-as-judge ratings, or curated case studies -- all subjective, none falsifiable. We formalize historical backtesting as an alternative: a system generates questions from a corpus frozen at a historical cutoff, the questions are frozen before any access to later literature, and a temporally isolated future corpus then determines whether each question was subsequently answered, partially addressed, independently posed, or ignored, and whether its underlying premise was supported or refuted. The protocol is model-agnostic: any system that emits frozen questions can be scored. We release reproducible astronomy instances with temporally isolated corpora, frozen questions, auditable labels, four reference baselines, and a submission interface. Two findings result. First, evidence-structure-first generation outperforms LLM-only prompting: across a generator decomposition crossed with a four-cutoff stress test (2010-2024, 798 judged questions) whose last window postdates model training, LLM-only generation shows memorized relevance without specific foresight, while a generator using no model weights at all finds questions whose premises the future refutes in every era. Second, a seven-rater agreement study (two blinded human annotators, five judge models, 90 items) indicts the outcome taxonomy rather than the judge: two careful humans agree at kappa = 0.17, every judge model agrees with the professional annotator as well or better (0.17-0.26), and frontier models agree with one another at 0.60 -- certifying an LLM judge by model-model agreement would have overstated its reliability threefold. A prospective instance -- 200 questions frozen 2026-08-17, scored 2027-2030 -- is released so the central claims become contamination-free tests that time itself will grade.

Human evaluation is critical for validating the performance of text-to-image generative models, as this highly cognitive process requires deep comprehension of text and images. However, our survey of 37 recent papers reveals that many works rely solely on automatic measures (e.g., FID) or perform poorly described human evaluations that are not reliable or repeatable. This paper proposes a standardized and well-defined human evaluation protocol to facilitate verifiable and reproducible human evaluation in future works. In our pilot data collection, we experimentally show that the current automatic measures are incompatible with human perception in evaluating the performance of the text-to-image generation results. Furthermore, we provide insights for designing human evaluation experiments reliably and conclusively. Finally, we make several resources publicly available to the community to facilitate easy and fast implementations.

Generating step-by-step "how-to" procedures is a key LLM capability: how-to advice is commonly requested in chatbots, and step-by-step planning is critical for reasoning over complex tasks. Yet, measuring and improving procedural validity at scale on real-world tasks remains challenging and understudied. To address this, we introduce How2Everything, a scalable framework to evaluate and improve goal-conditioned procedure generation. Our framework includes How2Mine, which mines 351K procedures from 980K web pages across 14 topics and readily scales to larger corpora. From this pool we build How2Bench, a 7K-example evaluation set balanced across topics. To reliably score model outputs, we develop How2Score, an evaluation protocol that uses an LLM judge to detect whether a generation contains any critical failure that would prevent achieving the goal. For low-cost, reproducible evaluation, we distill a frontier model into an open 8B model, achieving 80.5% agreement with human annotators. How2Bench reveals clear scaling trends across model sizes and training stages, providing signal early in pretraining. Finally, RL using How2Score as a reward improves performance on How2Bench by >10 points across three models without systematic regressions on standard benchmarks, with gains robust to superficial source-document memorization or format compliance. Taken together, How2Everything shows how pretraining web data can support a closed loop of capability evaluation and improvement at scale.

Medical care follows complex clinical pathways that extend beyond isolated physician-patient encounters, emphasizing decision-making and transitions between different stages. Current benchmarks focusing on static exams or isolated dialogues inadequately evaluate large language models (LLMs) in dynamic clinical scenarios. We introduce CP-Env, a controllable agentic hospital environment designed to evaluate LLMs across end-to-end clinical pathways. CP-Env simulates a hospital ecosystem with patient and physician agents, constructing scenarios ranging from triage and specialist consultation to diagnostic testing and multidisciplinary team meetings for agent interaction. Following real hospital adaptive flow of healthcare, it enables branching, long-horizon task execution. We propose a three-tiered evaluation framework encompassing Clinical Efficacy, Process Competency, and Professional Ethics. Results reveal that most models struggle with pathway complexity, exhibiting hallucinations and losing critical diagnostic details. Interestingly, excessive reasoning steps can sometimes prove counterproductive, while top models tend to exhibit reduced tool dependency through internalized knowledge. CP-Env advances medical AI agents development through comprehensive end-to-end clinical evaluation. We provide the benchmark and evaluation tools for further research and development at https://github.com/SPIRAL-MED/CP_ENV.

Recent machine learning papers often report 1-2 percentage point improvements from a single run on a benchmark. These gains are highly sensitive to random seeds, data ordering, and implementation details, yet are rarely accompanied by uncertainty estimates or significance tests. It is therefore unclear when a reported +1-2% reflects a real algorithmic advance versus noise. We revisit this problem under realistic compute budgets, where only a few runs are affordable. We propose a simple, PC-friendly evaluation protocol based on paired multi-seed runs, bias-corrected and accelerated (BCa) bootstrap confidence intervals, and a sign-flip permutation test on per-seed deltas. The protocol is intentionally conservative and is meant as a guardrail against over-claiming. We instantiate it on CIFAR-10, CIFAR-10N, and AG News using synthetic no-improvement, small-gain, and medium-gain scenarios. Single runs and unpaired t-tests often suggest significant gains for 0.6-2.0 point improvements, especially on text. With only three seeds, our paired protocol never declares significance in these settings. We argue that such conservative evaluation is a safer default for small gains under tight budgets.

Artificial intelligence systems are increasingly deployed in biomedical research. However, current evaluation frameworks may inadequately assess their effectiveness as research collaborators. This rapid review examines benchmarking practices for AI systems in preclinical biomedical research. Three major databases and two preprint servers were searched from January 1, 2018 to October 31, 2025, identifying 14 benchmarks that assess AI capabilities in literature understanding, experimental design, and hypothesis generation. The results revealed that all current benchmarks assess isolated component capabilities, including data analysis quality, hypothesis validity, and experimental protocol design. However, authentic research collaboration requires integrated workflows spanning multiple sessions, with contextual memory, adaptive dialogue, and constraint propagation. This gap implies that systems excelling on component benchmarks may fail as practical research co-pilots. A process-oriented evaluation framework is proposed that addresses four critical dimensions absent from current benchmarks: dialogue quality, workflow orchestration, session continuity, and researcher experience. These dimensions are essential for evaluating AI systems as research co-pilots rather than as isolated task executors.

Frontier language models sometimes recognize that they are being evaluated and adjust their behavior, undermining validity of benchmark results. Yet the field studies it without a shared foundation, conflating properties of the evaluation with properties of the model, and detection with behavioral response. We ground evaluation awareness in social psychology, decomposing it into an environment component (how recognizable the task is) and a model component that separates recognition from propensity to act on it. We operationalize the environment component through eight categorized trigger factors, such as placeholder entities and grading-style output formats, and study recognition and behavior through chain-of-thought monitoring. Across nine frontier models and four benchmarks, recognition rates depend on the specific pairing of model and benchmark rather than on either in isolation. Recognition rarely leads to behavioral change, and when it does, the direction depends on the type of evaluation perceived. Models are also more sensitive to safety than capability evaluations, placing safety benchmark validity at greater risk. To study which factors each model is sensitive to and how they interact, we propose EvalAwareBench, a factor-controlled benchmark of 100 paired safety-capability tasks where each of the eight factors can be independently toggled, varying evaluative signals while holding the underlying request fixed. Through EvalAwareBench, we find that no single factor uniformly affects all models, but stacking factors progressively raises evaluation awareness across all of them. Our framework and EvalAwareBench provide the tools to measure, attribute, and mitigate evaluation awareness, pointing to behavioral consistency under recognition as a promising path forward.

Various algorithms for continual learning (CL) have been designed with the goal of effectively alleviating the trade-off between stability and plasticity during the CL process. To achieve this goal, tuning appropriate hyperparameters for each algorithm is essential. As an evaluation protocol, it has been common practice to train a CL algorithm using diverse hyperparameter values on a CL scenario constructed with a benchmark dataset. Subsequently, the best performance attained with the optimal hyperparameter value serves as the criterion for evaluating the CL algorithm. In this paper, we contend that this evaluation protocol is not only impractical but also incapable of effectively assessing the CL capability of a CL algorithm. Returning to the fundamental principles of model evaluation in machine learning, we propose an evaluation protocol that involves Hyperparameter Tuning and Evaluation phases. Those phases consist of different datasets but share the same CL scenario. In the Hyperparameter Tuning phase, each algorithm is iteratively trained with different hyperparameter values to find the optimal hyperparameter values. Subsequently, in the Evaluation phase, the optimal hyperparameter values is directly applied for training each algorithm, and their performance in the Evaluation phase serves as the criterion for evaluating them. Through experiments on CIFAR-100 and ImageNet-100 based on the proposed protocol in class-incremental learning, we not only observed that the existing evaluation method fail to properly assess the CL capability of each algorithm but also observe that some recently proposed state-of-the-art algorithms, which reported superior performance, actually exhibit inferior performance compared to the previous algorithm.

Eliciting detailed and actionable software requirements from end-users is a critical phase in the iterative development of a software product or application. To ensure the feedback collected is detailed and actionable, software teams can leverage the laddering interview technique. While effective for ensuring granular and actionable items from the software feedback, these interviews are subject to several limitations. They are traditionally a manual process associated with a time and financial burden, limiting scalability; interviewers must balance probing for depth while managing interviewee behavioral and cultural constraints. To address these limitations, we present LadderTeam, an open, reproducible framework that automates UX wireframe interviews using a dual-agent Large Language Model (LLM) architecture. An active interviewer agent executes one of three probing strategies (ACV, 5-Whys, and JTBD) to elicit actionable software requirements from usability feedback comments, while a concurrent background Judge agent evaluates probe-response pairs and triggers real-time guardrails to prevent topic drift. To rigorously evaluate LLM laddering without participant variance confounds, we introduce a controlled simulation methodology utilizing scripted ground-truth transcripts to isolate probe quality as the sole experimental variable. Across 216 interviews, LadderTeam achieved 99.1\% chain convergence and an 81.0\% ground-truth actionable response match (86.1\% reluctant personality, 75.9\% terse personality) with zero drift across all runs. All evaluation code, all transcripts, inputs, and a live demonstration platform will be open-sourced upon acceptance.

Large language models are increasingly deployed as autonomous agents executing multi-step workflows in real-world software environments. However, existing agent benchmarks suffer from three critical limitations: (1) trajectory-opaque grading that checks only final outputs, (2) underspecified safety and robustness evaluation, and (3) narrow modality coverage and interaction paradigms. We introduce Claw-Eval, an end-to-end evaluation suite addressing all three gaps. It comprises 300 human-verified tasks spanning 9 categories across three groups (general service orchestration, multimodal perception and generation, and multi-turn professional dialogue). Every agent action is recorded through three independent evidence channels (execution traces, audit logs, and environment snapshots), enabling trajectory-aware grading over 2,159 fine-grained rubric items. The scoring protocol evaluates Completion, Safety, and Robustness, reporting Average Score, Pass@k, and Pass^k across three trials to distinguish genuine capability from lucky outcomes. Experiments on 14 frontier models reveal that: (1) trajectory-opaque evaluation is systematically unreliable, missing 44% of safety violations and 13% of robustness failures that our hybrid pipeline catches; (2) controlled error injection primarily degrades consistency rather than peak capability, with Pass^3 dropping up to 24% while Pass@3 remains stable; (3) multimodal performance varies sharply, with most models performing poorer on video than on document or image, and no single model dominating across all modalities. Beyond benchmarking, Claw-Eval highlights actionable directions for agent development, shedding light on what it takes to build agents that are not only capable but reliably deployable.

Agent benchmarks increasingly evaluate repository editing, web research, terminal use, and long-horizon interaction. Their scores support capability claims only when the evaluation protocol keeps the intended capability necessary for success. Recent reward-hacking benchmarks and system reports show that agents can instead recover public solutions, read evaluation artifacts, infer generator structure, manipulate feedback, or benefit from invalid scoring paths; existing responses do not provide a common procedure for attributing these shortcuts and quantifying their effect across benchmarks. We formulate protocol validity and introduce HackDetect, a post-hoc audit that identifies an exposure, determines how the agent used it, and assesses whether the resulting score is misleading. We quantify score inflation with the Mislead gap, defined as the exploit score minus the intended score. We audit 2,385 traces across 15 agent benchmarks and find evidence of exposures and reward hacking in 67.0% of Frontier Science traces and 66.7% of AutoLab tasks. Across paired comparisons, we measure score inflation of 0.45-1.00, showing that benchmark reports should provide evidence that scores reflect the intended capability.

The Model Context Protocol (MCP), introduced by Anthropic in November 2024, defines a standardized interface for connecting large language models (LLMs) to external tools, data sources, and services. Within months of release, hundreds of community-built MCP servers appeared on GitHub, but no software-maintenance literature has yet described how the ecosystem is being structured in production. This industry experience paper catalogues five recurring MCP server architectural patterns observed across an enumerated corpus of fifteen independently developed servers (five production servers from the ANSYR voice AI platform plus ten public servers from the official MCP registry): Resource Gateway, Tool Orchestrator, Stateful Session Server, Proxy Aggregator, and Domain-Specific Adapter. Each pattern is described in the structured form of Gamma et al.: context, problem, solution, and consequences. We also document four anti-patterns and a set of cross-cutting concerns around authentication, versioning, and observability. The quantitative evaluation contributes three measurements: inter-rater reliability of the taxonomy across two independent LLM raters on 54 held-out servers (Cohen's kappa = 0.76), which also localizes three pattern-boundary ambiguities; transport overhead measured end-to-end on loopback and modeled for cross-host paths; and a tool-count study showing tool-selection accuracy drops below 90% between 10 and 15 tools per context for Claude Haiku 4.5 and between 20 and 30 tools for Sonnet 4. Code, corpus, and prompts are released as a replication package.

Recent advancements in visual generative models have enabled high-quality image and video generation, opening diverse applications. However, evaluating these models often demands sampling hundreds or thousands of images or videos, making the process computationally expensive, especially for diffusion-based models with inherently slow sampling. Moreover, existing evaluation methods rely on rigid pipelines that overlook specific user needs and provide numerical results without clear explanations. In contrast, humans can quickly form impressions of a model's capabilities by observing only a few samples. To mimic this, we propose the Evaluation Agent framework, which employs human-like strategies for efficient, dynamic, multi-round evaluations using only a few samples per round, while offering detailed, user-tailored analyses. It offers four key advantages: 1) efficiency, 2) promptable evaluation tailored to diverse user needs, 3) explainability beyond single numerical scores, and 4) scalability across various models and tools. Experiments show that Evaluation Agent reduces evaluation time to 10% of traditional methods while delivering comparable results. The Evaluation Agent framework is fully open-sourced to advance research in visual generative models and their efficient evaluation.

Pass@k is widely used to report the reasoning performance of LLMs, but it often produces unstable and potentially misleading rankings, especially when the number of trials (samples) is limited and computational resources are constrained. We present a principled Bayesian evaluation framework that replaces Pass@k and average accuracy over N trials (avg@N) with posterior estimates of a model's underlying success probability and credible intervals, yielding stable rankings and a transparent decision rule for differences. Evaluation outcomes are modeled as categorical (not just 0/1) with a Dirichlet prior, giving closed-form expressions for the posterior mean and uncertainty of any weighted rubric and enabling the use of prior evidence when appropriate. Theoretically, under a uniform prior, the Bayesian posterior mean is order-equivalent to average accuracy (Pass@1), explaining its empirical robustness while adding principled uncertainty. Empirically, in simulations with known ground-truth success rates and on AIME'24/'25, HMMT'25, and BrUMO'25, the posterior-based procedure achieves faster convergence and greater rank stability than Pass@k and recent variants, enabling reliable comparisons at far smaller sample counts. The framework clarifies when observed gaps are statistically meaningful (non-overlapping credible intervals) versus noise, and it naturally extends to graded, rubric-based evaluations. Together, these results recommend replacing Pass@k for LLM evaluation and ranking with a posterior-based, compute-efficient protocol that unifies binary and non-binary evaluation while making uncertainty explicit. Source code is available at https://github.com/mohsenhariri/scorio

Rubric-based evaluation has become a prevailing paradigm for evaluating instruction following in large language models (LLMs). Despite its widespread use, the reliability of these rubric-level evaluations remains unclear, calling for meta-evaluation. However, prior meta-evaluation efforts largely focus on the response level, failing to assess the fine-grained judgment accuracy that rubric-based evaluation relies on. To bridge this gap, we introduce RubricEval. Our benchmark features: (1) the first rubric-level meta-evaluation benchmark for instruction following, (2) diverse instructions and responses spanning multiple categories and model sources, and (3) a substantial set of 3,486 quality-controlled instances, along with Easy/Hard subsets that better differentiates judge performance. Our experiments reveal that rubric-level judging remains far from solved: even GPT-4o, a widely adopted judge in instruction-following benchmarks, achieves only 55.97% on Hard subset. Considering evaluation paradigm, rubric-level evaluation outperforms checklist-level, explicit reasoning improves accuracy, and both together reduce inter-judge variance. Through our established rubric taxonomy, we further identify common failure modes and offer actionable insights for reliable instruction-following evaluation.

Representation engineering methods have recently shown promise for enabling efficient steering of model behavior. However, evaluation pipelines for these methods have primarily relied on subjective demonstrations, instead of quantitative, objective metrics. We aim to take a step towards addressing this issue by advocating for four properties missing from current evaluations: (i) contexts sufficiently similar to downstream tasks should be used for assessing intervention quality; (ii) model likelihoods should be accounted for; (iii) evaluations should allow for standardized comparisons across different target behaviors; and (iv) baseline comparisons should be offered. We introduce an evaluation pipeline grounded in these criteria, offering both a quantitative and visual analysis of how effectively a given method works. We use this pipeline to evaluate two representation engineering methods on how effectively they can steer behaviors such as truthfulness and corrigibility, finding that some interventions are less effective than previously reported.

Ambient AI scribes draft clinical notes, and published audits find their dominant error is omission: information the encounter established that the note fails to record. The standard check is an LLM judge: a second model reads the note against the transcript and flags problems. We ask whether judges detect omissions. Public corpora cannot supply the answer key: their clinician reference notes and transcripts are materially discrepant. Our benchmark has 500 single-error note pairs from audited fact sheets, 298 with a named fact certainly absent and 202 added-or-altered controls. Across eight judge designs, paired discrimination (the flawed note below its clean twin, 0.5 a coin flip) reads 0.79-0.94 on added or altered content and 0.50-0.63 on omissions. On single notes, no design flags omissions reliably more often than perfect notes. Wording changes, voting and GEPA prompt optimisation move the operating point without creating usable detection. Restructuring the task recovers it: list the facts the transcript establishes, then check the note for each. Two methods reach it independently and trade off: a per-fact pipeline, and a GEPA-evolved prompt doing the same in one call. The pipeline's flags name the missing fact and its severity at 2.7% false alarms. The single call detects more (36.9% against 24.6%, p=0.002) at 6.2% false alarms and a tenth of the cost per note. A physician author validated 70 items and, where the two routes disagree, sided with the pipeline on 10 of 10 (p=0.002). A second clinician, not an author, graded the severity rubric blind and agrees to within a grade. On real vendor notes from a companion census no benchmark threshold transfers, but the re-calibrated single call detects more than the best of the eight at half its false-alarm rate. Omissions whose fact is restated elsewhere defeat both routes. We release the benchmark, prompts and judgements.

Recent advances in text-to-speech (TTS) technology have enabled systems to produce human-indistinguishable speech, bringing benefits across accessibility, content creation, and human-computer interaction. However, current evaluation practices are increasingly inadequate for capturing the full range of capabilities, limitations, and societal implications. This position paper introduces the concept of Responsible Evaluation and argues that it is essential and urgent for the next phase of TTS development, structured through three progressive levels: (1) ensuring the faithful and accurate reflection of a model's true capabilities, with more robust, discriminative, and comprehensive objective and subjective scoring methodologies; (2) enabling comparability, standardization, and transferability through standardized benchmarks, transparent reporting, and transferable evaluation metrics; and (3) assessing and mitigating ethical risks associated with forgery, misuse, privacy violations, and security vulnerabilities. Through this concept, we critically examine current evaluation practices, identify systemic shortcomings, and propose actionable recommendations. We hope this concept of Responsible Evaluation will foster more trustworthy and reliable TTS technology and guide its development toward ethically sound and societally beneficial applications.

As the use of text-to-image generative models increases, so does the adoption of automatic benchmarking methods used in their evaluation. However, while metrics and datasets abound, there are few unified benchmarking libraries that provide a framework for performing evaluations across many datasets and metrics. Furthermore, the rapid introduction of increasingly robust benchmarking methods requires that evaluation libraries remain flexible to new datasets and metrics. Finally, there remains a gap in synthesizing evaluations in order to deliver actionable takeaways about model performance. To enable unified, flexible, and actionable evaluations, we introduce EvalGIM (pronounced ''EvalGym''), a library for evaluating generative image models. EvalGIM contains broad support for datasets and metrics used to measure quality, diversity, and consistency of text-to-image generative models. In addition, EvalGIM is designed with flexibility for user customization as a top priority and contains a structure that allows plug-and-play additions of new datasets and metrics. To enable actionable evaluation insights, we introduce ''Evaluation Exercises'' that highlight takeaways for specific evaluation questions. The Evaluation Exercises contain easy-to-use and reproducible implementations of two state-of-the-art evaluation methods of text-to-image generative models: consistency-diversity-realism Pareto Fronts and disaggregated measurements of performance disparities across groups. EvalGIM also contains Evaluation Exercises that introduce two new analysis methods for text-to-image generative models: robustness analyses of model rankings and balanced evaluations across different prompt styles. We encourage text-to-image model exploration with EvalGIM and invite contributions at https://github.com/facebookresearch/EvalGIM/.

Randomized controlled trials (RCTs) are the benchmark for causal inference, yet field implementation can deviate. We here present a remote audit - a design-based, preregistrable diagnostic that uses only pre-treatment satellite imagery to test whether assignment is independent of local conditions. The conditional randomization test of the remote audit evaluates whether treatment assignment is more predictable from pre-treatment satellite features than expected under the experiment's registered mechanism, providing a finite-sample valid, design-based diagnostic that requires no parametric assumptions. The procedure is finite-sample valid, honors blocks and clusters, and controls multiplicity across image models and resolutions via a max-statistic. We illustrate with two RCTs: Uganda's Youth Opportunities Program, where the audit corroborates randomization and flags selection and missing-data risks; and a school-based trial in Bangladesh, where assignment is highly predictable from pre-treatment features relative to the stated design, consistent with independent concerns about irregularities. Remote audits complement balance tests, lower early-stage costs, and enable rapid design checks when baseline surveys are expensive or infeasible.

Recent progress in deep research systems has been impressive, but evaluation still lags behind real user needs. Existing benchmarks predominantly assess final reports using fixed rubrics, failing to evaluate the underlying research process. Most also offer limited multimodal coverage, rely on synthetic tasks that do not reflect real-world query complexity, and cannot be refreshed as knowledge evolves. To address these gaps, we introduce MiroEval, a benchmark and evaluation framework for deep research systems. The benchmark comprises 100 tasks (70 text-only, 30 multimodal), all grounded in real user needs and constructed via a dual-path pipeline that supports periodic updates, enabling a live and evolving setting. The proposed evaluation suite assesses deep research systems along three complementary dimensions: adaptive synthesis quality evaluation with task-specific rubrics, agentic factuality verification via active retrieval and reasoning over both web sources and multimodal attachments, and process-centric evaluation audits how the system searches, reasons, and refines throughout its investigation. Evaluation across 13 systems yields three principal findings: the three evaluation dimensions capture complementary aspects of system capability, with each revealing distinct strengths and weaknesses across systems; process quality serves as a reliable predictor of overall outcome while revealing weaknesses invisible to output-level metrics; and multimodal tasks pose substantially greater challenges, with most systems declining by 3 to 10 points. The MiroThinker series achieves the most balanced performance, with MiroThinker-H1 ranking the highest overall in both settings. Human verification and robustness results confirm the reliability of the benchmark and evaluation framework. MiroEval provides a holistic diagnostic tool for the next generation of deep research agents.

Large language models (LLMs) are increasingly proposed as clinical agents, yet static, single-turn benchmarks cannot capture how a model dynamically delivers care across an encounter: gathering information, planning treatment, and adapting longitudinal management across successive patient states. Medical education has long addressed an analogous challenge through standardized patients (SPs): trained actors who consistently portray clinical cases, enabling realistic practice and objective, scripted assessment. Here we introduce MedSP1000, an SP-derived interactive benchmark for clinical-agent evaluation, including 1,638 SP cases with 24,602 trajectory-level peer-reviewed rubrics. MedSP1000 converts peer-reviewed SP teaching cases into executable scenarios with defined SP case scripts, clinical environment contexts, and human-validated structured rubric. In each simulation evaluation run, a clinical agent interacts in closed loop with a patient agent and an environment controller, and its behaviour is scored throughout the encounter against expert criteria specified in the original materials. Applying MedSP1000 to a range of general-purpose and medically specialized LLMs, we find that performance on static benchmarks does not reliably translate to such educational scenarios. The best-performing model, GPT-5.5, completes only 60.4% of expert-defined rubric items, whereas the strongest medically specialized model reaches 40.0%; increasing test-time compute produces no measurable gain. These results suggest that current LLMs, including agentic systems tuned for medicine, are not yet reliable enough to be safely integrated into actual clinical practice. More broadly, MedSP1000 shows how process-level, SP-style evaluation can reveal clinically relevant failure modes that single-turn benchmarks miss.

In this study, we introduce FEET, a standardized protocol designed to guide the development and benchmarking of foundation models. While numerous benchmark datasets exist for evaluating these models, we propose a structured evaluation protocol across three distinct scenarios to gain a comprehensive understanding of their practical performance. We define three primary use cases: frozen embeddings, few-shot embeddings, and fully fine-tuned embeddings. Each scenario is detailed and illustrated through two case studies: one in sentiment analysis and another in the medical domain, demonstrating how these evaluations provide a thorough assessment of foundation models' effectiveness in research applications. We recommend this protocol as a standard for future research aimed at advancing representation learning models.

AI evaluation is undergoing a structural change. Large language models (LLMs) are increasingly deployed as systems that act over time through tools, environments, users, and other agents, while many evaluation practices still inherit assumptions from response-centered benchmarks (e.g., fixed inputs, isolated outputs, and outcome judgments that can be made from a single response). The field has begun to build interactive benchmarks, but the resulting landscape is fragmented: benchmarks differ in what interaction artifacts they admit, how trajectories are scored, and what claims their results support. This position paper argues that interactive evaluation should be treated as a principled evaluation paradigm, not merely a new family of agent benchmarks. Simply adopting previous evaluation paradigms does not suffice. We define evaluation as an autonomous mapping from evidence to judgments, and show that interactive evaluation changes both sides of this mapping: the evidence becomes interaction-generated trajectories, while the evaluation procedure must assess process, recoverability, coordination, robustness, and system-level performance. Building on this definition, we propose a two-axis taxonomy, derive design principles and reporting standards, examine representative scenarios, and analyze how longstanding evaluation challenges reappear at the trajectory level.

The computation necessary for training Transformer-based language models has skyrocketed in recent years. This trend has motivated research on efficient training algorithms designed to improve training, validation, and downstream performance faster than standard training. In this work, we revisit three categories of such algorithms: dynamic architectures (layer stacking, layer dropping), batch selection (selective backprop, RHO loss), and efficient optimizers (Lion, Sophia). When pre-training BERT and T5 with a fixed computation budget using such methods, we find that their training, validation, and downstream gains vanish compared to a baseline with a fully-decayed learning rate. We define an evaluation protocol that enables computation to be done on arbitrary machines by mapping all computation time to a reference machine which we call reference system time. We discuss the limitations of our proposed protocol and release our code to encourage rigorous research in efficient training procedures: https://github.com/JeanKaddour/NoTrainNoGain.

We introduce AnyTool, a large language model agent designed to revolutionize the utilization of a vast array of tools in addressing user queries. We utilize over 16,000 APIs from Rapid API, operating under the assumption that a subset of these APIs could potentially resolve the queries. AnyTool primarily incorporates three elements: an API retriever with a hierarchical structure, a solver aimed at resolving user queries using a selected set of API candidates, and a self-reflection mechanism, which re-activates AnyTool if the initial solution proves impracticable. AnyTool is powered by the function calling feature of GPT-4, eliminating the need for training external modules. We also revisit the evaluation protocol introduced by previous works and identify a limitation in this protocol that leads to an artificially high pass rate. By revising the evaluation protocol to better reflect practical application scenarios, we introduce an additional benchmark, termed AnyToolBench. Experiments across various datasets demonstrate the superiority of our AnyTool over strong baselines such as ToolLLM and a GPT-4 variant tailored for tool utilization. For instance, AnyTool outperforms ToolLLM by +35.4% in terms of average pass rate on ToolBench. Code will be available at https://github.com/dyabel/AnyTool.

The inherent rigidity of fixed-size benchmarks makes them an inefficient tool for model evaluation. Diverse evaluation objectives, including model ranking, model selection and testing throughout development, demand varying levels of statistical power. The mismatch between fixed sample sizes and these diverse needs results in either excessive computational cost or compromised reliability - a critical concern for model evaluation. To overcome these limitations, we call for adoption of sequential testing in our field. We provide an adaptive evaluation framework, that provides a principled way to navigate the trade-off between efficiency and reliability in model evaluation. Our framework combines the established statistical paradigm of sequential testing with stopping criteria tailored to common evaluation needs such as diminishing returns detection, and minimum detectable effect size. We demonstrate its ability to adaptively manage the efficiency-reliability trade-off on the Open VLM Leaderboard, including, for example, a 80% reduction in computational cost compared to fixed-size evaluation (with a 2.5-point CI width allowance) while maintaining statistical significance.

The acceleration of Large Language Models (LLMs) research has opened up new possibilities for evaluating generated texts. They serve as scalable and economical evaluators, but the question of how reliable these evaluators are has emerged as a crucial research question. Prior research efforts in the meta-evaluation of LLMs as judges limit the prompting of an LLM to a single use to obtain a final evaluation decision. They then compute the agreement between LLMs' outputs and human labels. This lacks interpretability in understanding the evaluation capability of LLMs. In light of this challenge, we propose Decompose and Aggregate, which breaks down the evaluation process into different stages based on pedagogical practices. Our experiments illustrate that it not only provides a more interpretable window for how well LLMs evaluate, but also leads to improvements up to 39.6% for different LLMs on a variety of meta-evaluation benchmarks.

Background: Agent skills are increasingly deployed as modular, reusable capability units in AI agent systems. Medical research agent skills require safeguards beyond general-purpose evaluation, including scientific integrity, methodological validity, reproducibility, and boundary safety. This study developed and preliminarily evaluated a domain-specific audit framework for medical research agent skills, with a focus on reliability against expert review. Methods: We developed MedSkillAudit (skill-auditor@1.0), a layered framework assessing skill release readiness before deployment. We evaluated 75 skills across five medical research categories (15 per category). Two experts independently assigned a quality score (0-100), an ordinal release disposition (Production Ready / Limited Release / Beta Only / Reject), and a high-risk failure flag. System-expert agreement was quantified using ICC(2,1) and linearly weighted Cohen's kappa, benchmarked against the human inter-rater baseline. Results: The mean consensus quality score was 72.4 (SD = 13.0); 57.3% of skills fell below the Limited Release threshold. MedSkillAudit achieved ICC(2,1) = 0.449 (95% CI: 0.250-0.610), exceeding the human inter-rater ICC of 0.300. System-consensus score divergence (SD = 9.5) was smaller than inter-expert divergence (SD = 12.4), with no directional bias (Wilcoxon p = 0.613). Protocol Design showed the strongest category-level agreement (ICC = 0.551); Academic Writing showed a negative ICC (-0.567), reflecting a structural rubric-expert mismatch. Conclusions: Domain-specific pre-deployment audit may provide a practical foundation for governing medical research agent skills, complementing general-purpose quality checks with structured audit workflows tailored to scientific use cases.

Clinical diagnosis is critical in medical practice, typically requiring a continuous and evolving process that includes primary diagnosis, differential diagnosis, and final diagnosis. However, most existing clinical diagnostic tasks are single-step processes, which does not align with the complex multi-step diagnostic procedures found in real-world clinical settings. In this paper, we propose a multi-step diagnostic task and annotate a clinical diagnostic dataset (MSDiagnosis). This dataset includes primary diagnosis, differential diagnosis, and final diagnosis questions. Additionally, we propose a novel and effective framework. This framework combines forward inference, backward inference, reflection, and refinement, enabling the LLM to self-evaluate and adjust its diagnostic results. To assess the effectiveness of our proposed method, we design and conduct extensive experiments. The experimental results demonstrate the effectiveness of the proposed method. We also provide a comprehensive experimental analysis and suggest future research directions for this task.

Front-end web code has become a core product surface for every frontier LLM release, yet evaluating these interactive applications at development speed remains costly because human-judged leaderboards like Arena do not scale. Existing automated proxies typically lean on reference implementations, test suites, or rigid checklists, and tend to miss the reasoned synthesis a human reviewer performs over a live session. We articulate a new evaluation regime that is simultaneously reference-free, autonomously driven, and holistically reasoned, and instantiate it through two artifacts. \dataname is an 11-domain, 54-leaf, 1,000-query WebDev benchmark spanning both static-presentation and interactive-application tasks, balanced across three difficulty tiers and three target-language groups, with briefs rewritten to resist recall from circulated prompts. \framename, grounded in Flavell's metacognitive monitoring, separates evidence accumulation from judgment across three stages: Static Perception forms a first impression from passive observation; Agent-Driven Interaction explores the application autonomously while capturing continuous screen video, audio, and per-step screenshots; Dynamic Scoring issues holistic functionality and aesthetics verdicts with structured failure attribution only after the evidence chain is complete. On \dataname, \framename aligns closely with expert human ratings while surfacing substantial headroom across 13 frontier LLMs on interactive web generation. \noindenthttps://anonymous.4open.science/r/Cookie-3CE/

ReAct-style agents for search-intensive, multi-step reasoning tasks rely largely on their own internal judgment to decide what evidence to seek, which reasoning or action step to take next, and when to stop, often producing shallow, redundant, or poorly targeted trajectories. Prior work has explored rubrics as external quality signals, but existing uses are mostly evaluative rather than action-guiding: rubrics typically serve as training-time rewards or post-hoc evaluators of completed outputs, and in deep-research settings they are often coarse-grained and report-level rather than step-level. We introduce Co-ReAct, a rubric-guided action-selection framework that uses rubrics as step-level guidance during inference. At each decision step, Co-ReAct injects a rubric into the agent's context to guide the next Reason-or-Act decision, specifying what the agent should target in evidence seeking, search, reasoning, or self-evaluation. To make this guidance reliable, we train a dedicated rubric generator with GRPO. Unlike prior pairwise or binary preference formulations, our objective optimizes a list-wise Spearman rank-correlation reward against multi-judge expert consensus rankings, encouraging rubrics that are discriminative rather than merely plausible. On DeepResearchBench and SQA-CS-V2, Co-ReAct consistently improves over ReAct and representative test-time compute baselines across search agents built on both 8B/14B open-source and frontier closed-source base models. The trained rubric generator can also serve as a drop-in component that improves these baselines without changing their underlying decision mechanisms. Our code is publicly available at https://github.com/ZBWpro/Co-ReAct.

As language models (LMs) evolve from chat assistants to long-horizon agents capable of multi-step reasoning and tool use, existing benchmarks remain largely confined to structured or exam-style tasks that fall short of real-world professional demands. To this end, we introduce \OneMillion-Bench OneMillion-Bench, a benchmark of 400 expert-curated tasks spanning Law, Finance, Industry, Healthcare, and Natural Science, built to evaluate agents across economically consequential scenarios. Unlike prior work, the benchmark requires retrieving authoritative sources, resolving conflicting evidence, applying domain-specific rules, and making constraint decisions, where correctness depends as much on the reasoning process as the final answer. We adopt a rubric-based evaluation protocol scoring factual accuracy, logical coherence, practical feasibility, and professional compliance, focused on expert-level problems to ensure meaningful differentiation across agents. Together, \$OneMillion-Bench provides a unified testbed for assessing agentic reliability, professional depth, and practical readiness in domain-intensive scenarios.

The evaluation paradigm of LLM-as-judge gains popularity due to its significant reduction in human labor and time costs. This approach utilizes one or more large language models (LLMs) to assess the quality of outputs from other LLMs. However, existing methods rely on generic scoring rubrics that fail to consider the specificities of each question and its problem-solving process, compromising precision and stability in assessments. Inspired by human examination scoring processes, we propose a new evaluation paradigm based on self-adaptive rubrics. Specifically, we create detailed scoring rubrics for each question, capturing the primary and secondary criteria in a structured format of scoring and deduction points that mimic a human evaluator's analytical process. Building on this paradigm, we further develop a novel benchmark called SedarEval, which covers a range of domains including long-tail knowledge, mathematics, coding, and logical reasoning. SedarEval consists of 1,000 meticulously crafted questions, each with its own self-adaptive rubric. To further streamline the evaluation, we train a specialized evaluator language model (evaluator LM) to supplant human graders. Using the same training data, our evaluator LM achieves a higher concordance rate with human grading results than other paradigms, including GPT-4, highlighting the superiority and efficiency of our approach. We release our dataset at https://github.com/wwn1233/sedareval.

A Mamba state-space model trained only for next-step prediction appears to recover Granger-causal structure through a simple readout S = |W_{out} W_{in}|, with early experiments suggesting the phenomenon generalized across architectures and benefited from interventional data at p < 10^{-5}. We package the protocol used to test that claim -- standardized synthetic generators (VAR/Lorenz/CauseMe-style), three intervention semantics (do(X=c), soft-noise, random-forcing), edge-provenance cards on three real datasets, and size-matched control arms -- as a reusable falsification benchmark, and walk the claim through it in five stages. The method-level claim does not survive: (i) a plain linear bottleneck does as well or better; (ii) tuned Lasso beats the bottleneck on synthetic CauseMe-style benchmarks, and on Lorenz-96 (the only real benchmark with unambiguous ground truth) classical PCMCI and Granger lead a tight cluster in which the bottleneck trails; (iii) the headline intervention advantage is roughly 60% a sample-size confound, and the residual disappears under standard do(X=c) interventions, surviving only under a non-standard random-forcing scheme; (iv) even that residual reproduces, with a larger effect, in classical bivariate Granger -- the effect is method-agnostic. What survives is a narrow characterization result; the benchmark is the lasting artifact, and each stage above is one of its control arms.

AI evaluation results are produced at scale but reported inconsistently across leaderboards, model cards, benchmark papers, and company blogs. The cost is interpretive: readers cannot reliably compare results across sources, identify what a report omits, or trace an aggregate claim to its underlying evidence. Recent efforts address isolated components but leave three gaps: they cover only narrow slices of the evaluation lifecycle and do not compose into a single interpretable record; they specify static representations that do not differentiate the questions different stakeholders bring to the same evidence; and they remain proposals on paper, lacking the extraction infrastructure required for adoption at scale. We present , an operational reporting layer that composes benchmark metadata, evaluation run data, and model metadata into a unified record. We (1) derive a reporting schema from a structured review of 52 papers and 10 stakeholder interviews, (2) implement four interpretive signals (reproducibility, documentation completeness, provenance and risk, and score comparability), rendered through reader modes calibrated to research and non-research audiences, and (3) deploy a monitoring tool that applies across 5,816 models, 635 benchmarks, and 101,843 results, surfacing systematic gaps in current reporting practice.

AI systems increasingly participate in their own improvement: revising their outputs, adapting their own harnesses during deployment, training on data they generate, and, increasingly, conducting AI research itself. This literature is described under a vocabulary ("self-refine," "self-reward," "self-play," "self-evolve") that conflates fundamentally different ambitions. We survey 1,250 arXiv papers (2024-2026) along two axes: what the system improves -- its behavior in deployment, its policy through training, its evaluator, or the research process itself -- and the degree of loop closure (human-in-the-loop to fully closed). The taxonomy separates bounded self-refinement -- convergent, evaluable, and already industrial practice -- from open-ended recursive self-improvement (RSI), which remains bounded by grounding requirements, collapse dynamics, and compute constraints on every measured axis. Its distinctive feature is a dedicated category for self-evaluation: every improvement loop is a claim that some signal can substitute for human judgment. We survey the evaluator design space -- judges, process reward models, verifiers, rubrics, meta-evaluation -- order the signals into a verification hierarchy from formal verifiers (strongest) to intrinsic self-assessment (weakest), and observe that demonstrated self-improvement strength tracks this hierarchy, that its failure modes (self-confirming loops, model collapse, diversity collapse) follow from its violations, and that the "research direction-setting" bottleneck keeping humans in the loop sits at the top of that hierarchy. We connect the technical literature to the theory of RSI limits and to the safety and governance questions raised by frontier-lab accounts of closing the loop, and identify governance-grade measurement of self-improvement as the field's most underpopulated niche.

We study auto research as a closed empirical loop driven by external measurement. Each submitted trial carries a hypothesis, an executable code edit, an evaluator-owned outcome, and feedback that shapes the next proposal. The output is not a generated paper or a single model checkpoint, but an auditable trajectory of proposals, code diffs, experiments, scores, and failure labels. We instantiate this loop with specialist agents that partition recipe surfaces and share measured lineage across trials. The central empirical finding is that lineage feedback lets agents turn evaluator outcomes, including crashes, budget overruns, size failures, and accuracy-gate misses, into later program-level recipe edits rather than one-shot suggestions. Across 1,197 headline-run trials plus 600 Parameter Golf control trials after one-time setup and launch, humans did not choose proposals, edit recipes, override scores, or repair failed trials during the search. In the three headline runs, the same submitted-trial loop reduces Parameter Golf validation bpb by 0.81%, raises NanoChat-D12 CORE by 38.7%, and reduces CIFAR-10 Airbench96 wallclock by 4.59%, with each task measured by its own external evaluator and legality checks. The trace includes a strict architecture-domain audit of 157 headline-run submissions and program rewrites such as a NanoChat attention-kernel path change. Within this scope the loop autonomously writes code, submits experiments, absorbs feedback, applies and combines known techniques inside each environment, and improves public starting recipes.

Foundation models are increasingly central to high-stakes AI systems, and governance frameworks now depend on evaluations to assess their risks and capabilities. Although general capability evaluations are widespread, social impact assessments covering bias, fairness, privacy, environmental costs, and labor practices remain uneven across the AI ecosystem. To characterize this landscape, we conduct the first comprehensive analysis of both first-party and third-party social impact evaluation reporting across a wide range of model developers. Our study examines 186 first-party release reports and 183 post-release evaluation sources, and complements this quantitative analysis with interviews of model developers. We find a clear division of evaluation labor: first-party reporting is sparse, often superficial, and has declined over time in key areas such as environmental impact and bias, while third-party evaluators including academic researchers, nonprofits, and independent organizations provide broader and more rigorous coverage of bias, harmful content, and performance disparities. However, this complementarity has limits. Only model developers can authoritatively report on data provenance, content moderation labor, financial costs, and training infrastructure, yet interviews reveal that these disclosures are often deprioritized unless tied to product adoption or regulatory compliance. Our findings indicate that current evaluation practices leave major gaps in assessing AI's societal impacts, highlighting the urgent need for policies that promote developer transparency, strengthen independent evaluation ecosystems, and create shared infrastructure to aggregate and compare third-party evaluations in a consistent and accessible way.

The evaluation of generative models for natural image tasks has been extensively studied. Similar protocols and metrics are used in cases with unique particularities, such as Handwriting Generation, even if they might not be completely appropriate. In this work, we introduce three measures tailored for HTG evaluation, HTG_{HTR} , HTG_{style} , and HTG_{OOV} , and argue that they are more expedient to evaluate the quality of generated handwritten images. The metrics rely on the recognition error/accuracy of Handwriting Text Recognition and Writer Identification models and emphasize writing style, textual content, and diversity as the main aspects that adhere to the content of handwritten images. We conduct comprehensive experiments on the IAM handwriting database, showcasing that widely used metrics such as FID fail to properly quantify the diversity and the practical utility of generated handwriting samples. Our findings show that our metrics are richer in information and underscore the necessity of standardized evaluation protocols in HTG. The proposed metrics provide a more robust and informative protocol for assessing HTG quality, contributing to improved performance in HTR. Code for the evaluation protocol is available at: https://github.com/koninik/HTG_evaluation.

There have been multiple calls for investments in the development of AI standards that both preserve the transformative potential and minimize the risks of AI. The goals of AI standards, particularly with respect to AI data, performance, and governance, are to promote innovation and public trust in systems that use AI. However, there is a lack of a formal or shared method to measure the impact of these standardization activities on the goals of innovation and trust. This concept paper proposes an analytical approach that could inform the evaluation of the impact of AI standards. The proposed approach could be used to measure, assess, and eventually evaluate the extent to which AI standards achieve their stated goals, since most Standards Development Organizationss do not track the impact of their standards once completed. It is intended to stimulate discussions with a wide variety of stakeholders, including academia and the standards community, about the potential for the approach to evaluate the effectiveness, utility, and relative value of AI standards. The document draws on successful and well-tested evaluation frameworks, tools, and metrics that are used for monitoring and assessing the effect of programmatic interventions in other domains to describe a possible approach. It begins by describing the context within which an evaluation would be designed, and then introduces a standard evaluation framework. These sections are followed by a description of what outputs and outcomes might result from the adoption and implementation of AI standards and the process whereby those AI standards are developed . Subsequent sections provide an overview of how the effectiveness of AI standards might be assessed and a conclusion.

Academic survey writing, which distills vast literature into a coherent and insightful narrative, remains a labor-intensive and intellectually demanding task. While recent approaches, such as general DeepResearch agents and survey-specialized methods, can generate surveys automatically (a.k.a. LLM4Survey), their outputs often fall short of human standards and there lacks a rigorous, reader-aligned benchmark for thoroughly revealing their deficiencies. To fill the gap, we propose a fine-grained, quiz-driven evaluation framework SurveyBench, featuring (1) typical survey topics source from recent 11,343 arXiv papers and corresponding 4,947 high-quality surveys; (2) a multifaceted metric hierarchy that assesses the outline quality (e.g., coverage breadth, logical coherence), content quality (e.g., synthesis granularity, clarity of insights), and non-textual richness; and (3) a dual-mode evaluation protocol that includes content-based and quiz-based answerability tests, explicitly aligned with readers' informational needs. Results show SurveyBench effectively challenges existing LLM4Survey approaches (e.g., on average 21% lower than human in content-based evaluation).

Recent advances have made long-form report-generating systems widely available. This has prompted evaluation frameworks that use LLM-as-judge protocols and claim verification, along with meta-evaluation frameworks that seek to validate these methods. Many of the meta-evaluations estimate an evaluation quality's by comparing its assessments against human pairwise preferences. Prior work, however, suggests that human pairwise preference may be overly simplistic and can fail to capture nuances of expert expectations. We conduct a case study in meta-evaluation for long-form QA benchmarks using ScholarQA-CS2, a benchmark designed for assessing retrieval-augmented deep-research QA in the scientific domain. We comprehensively validate the benchmark through human pairwise preference judgments, then critically examine the strengths, weaknesses, and confounders of this approach. We show that pairwise preference rankings are best suited for system-level evaluation, while explicit metric-wise annotations and expert annotators are critical for reliable metric-level assessment, with subjectivity remaining a key challenge. Based on our findings, we offer practical guidelines for designing future meta-evaluations that better align evaluation methods, annotator expertise, and reporting practices. By surfacing these methodological challenges, we aim to advance evaluation standards for deep-research systems.

The growing capabilities of large language models (LLMs) in instruction-following and context-understanding lead to the era of agents with numerous applications. Among these, task planning agents have become especially prominent in realistic scenarios involving complex internal pipelines, such as context understanding, tool management, and response generation. However, existing benchmarks predominantly evaluate agent performance based on task completion as a proxy for overall effectiveness. We hypothesize that merely improving task completion is misaligned with maximizing user satisfaction, as users interact with the entire agentic process and not only the end result. To address this gap, we propose PIPA, a unified evaluation protocol that conceptualizes the behavioral process of interactive task planning agents within a partially observable Markov Decision Process (POMDP) paradigm. The proposed protocol offers a comprehensive assessment of agent performance through a set of atomic evaluation criteria, allowing researchers and practitioners to diagnose specific strengths and weaknesses within the agent's decision-making pipeline. Our analyses show that agents excel in different behavioral stages, with user satisfaction shaped by both outcomes and intermediate behaviors. We also highlight future directions, including systems that leverage multiple agents and the limitations of user simulators in task planning.

Multi-model routing has evolved from an engineering technique into essential infrastructure, yet existing work lacks a systematic, reproducible benchmark for evaluating vision-language models (VLMs). We present VL-RouterBench to assess the overall capability of VLM routing systems systematically. The benchmark is grounded in raw inference and scoring logs from VLMs and constructs quality and cost matrices over sample-model pairs. In scale, VL-RouterBench covers 14 datasets across 3 task groups, totaling 30,540 samples, and includes 15 open-source models and 2 API models, yielding 519,180 sample-model pairs and a total input-output token volume of 34,494,977. The evaluation protocol jointly measures average accuracy, average cost, and throughput, and builds a ranking score from the harmonic mean of normalized cost and accuracy to enable comparison across router configurations and cost budgets. On this benchmark, we evaluate 10 routing methods and baselines and observe a significant routability gain, while the best current routers still show a clear gap to the ideal Oracle, indicating considerable room for improvement in router architecture through finer visual cues and modeling of textual structure. We will open-source the complete data construction and evaluation toolchain to promote comparability, reproducibility, and practical deployment in multimodal routing research.

Deep research agents increasingly automate complex information-seeking tasks, producing evidence-grounded reports via multi-step reasoning, tool use, and synthesis. Their growing role demands scalable, reliable evaluation, positioning LLM-as-judge as a supervision paradigm for assessing factual accuracy, evidence use, and reasoning quality. Yet the reliability of these judges for deep research agents remains poorly understood, posing a critical meta-evaluation problem: before deploying LLM judges to supervise research agents, we must first evaluate the judges themselves. Existing meta-evaluations fall short in two ways: (1) reliance on coarse, subjective human-preference agreement; (2) focus on instruction-following or verifiable tasks, leaving open-ended agent executions unexplored. To address these gaps, we introduce REFLECT (REliable Fine-grained LLM judge Evaluation via Controlled inTervention), a meta-evaluation benchmark targeting fine-grained failure detection in agentic environments. REFLECT defines a detailed taxonomy of process- and outcome-level failure modes, instantiated by performing controlled and localized interventions on quality-screened agent execution traces. This yields verifiable, comprehensive, and fine-grained instances for validating the judge models. Our experiments show that current LLM judges remain unreliable: even the best-performing models achieve overall accuracies below 55% across reasoning, tool-use, and report-quality failures, with especially poor performance on evidence verification. Together, our taxonomy and findings expose systematic judge limitations, reveal tradeoffs in cost and reliability, and offer actionable guidance for building more reliable evaluation pipelines for deep research agents.

The validity of AI safety evaluations depends on models behaving consistently across controlled and deployment settings. Prior work has identified test-time contextual cues, such as hypothetical scenarios, as a source of verbalized evaluation awareness and subsequent behavioral shift. In this paper, we investigate a potential explanation of this phenomenon: evaluation meta-knowledge, defined as parametric knowledge about the structural traits that characterize evaluations. Similar to dataset contamination, where benchmark exposure leads to higher performance through memorization, we hypothesize that models trained on texts describing evaluation practices may implicitly learn to recognize and respond to evaluation-like contexts, for instance, through exposure to scientific articles or social media posts about AI benchmarking. To test this, we fine-tune models on synthetic documents describing evaluation traits such as verifiable structures or moral dilemmas. Evaluating this fine-tuned model on six safety benchmarks, we find that it is significantly safer than the base model and control model. This behavioral shift persists even when restricting the analysis to responses lacking explicit verbalization of evaluation awareness. Our results demonstrate that evaluation meta-knowledge may inflate safety benchmark performance, introducing a novel confounder that is independent of explicit memorization or verbalized evaluation awareness, thus, challenging to detect. These findings have important implications for the design and interpretation of AI safety evaluations. Our code and models are available at https://github.com/compass-group-tue/arxiv2026_evaluation_meta_knowledge.

AI pentesting agents are increasingly credible as offensive security systems, but current benchmarks still provide limited guidance on which will perform best in real-world targets. Existing evaluation protocols assess and optimize for predefined goals such as capture-the-flag, remote code execution, exploit reproduction, or trajectory similarity, in simplified or narrow settings. These tools are valuable for measuring bounded capabilities, yet they do not adequately capture the complexity, open-ended exploration, and strategic decision-making required in realistic pentesting. In this paper, we present a practical evaluation protocol that shifts assessment from task completion to validated vulnerability discovery, allowing evaluation in sufficiently complex targets spanning multiple attack surfaces and vulnerability classes. The protocol combines structured ground-truth with LLM-based semantic matching to identify vulnerabilities, bipartite resolution to score findings under realistic ambiguity, continuous ground-truth maintenance, repeated and cumulative evaluation of stochastic agents, efficiency metrics, and reduced-suite selection for sustainable experimentation. This protocol extends the state of the art by enabling a more realistic, operationally informative comparison of AI pentesting agents. To enable reproducibility, we also release expert-annotated ground truth and code for the proposed evaluation protocol: https://github.com/ethiack/ethibench.

In recent years, Text-to-Image (T2I) models have been extensively studied, especially with the emergence of diffusion models that achieve state-of-the-art results on T2I synthesis tasks. However, existing benchmarks heavily rely on subjective human evaluation, limiting their ability to holistically assess the model's capabilities. Furthermore, there is a significant gap between efforts in developing new T2I architectures and those in evaluation. To address this, we introduce HRS-Bench, a concrete evaluation benchmark for T2I models that is Holistic, Reliable, and Scalable. Unlike existing bench-marks that focus on limited aspects, HRS-Bench measures 13 skills that can be categorized into five major categories: accuracy, robustness, generalization, fairness, and bias. In addition, HRS-Bench covers 50 scenarios, including fashion, animals, transportation, food, and clothes. We evaluate nine recent large-scale T2I models using metrics that cover a wide range of skills. A human evaluation aligned with 95% of our evaluations on average was conducted to probe the effectiveness of HRS-Bench. Our experiments demonstrate that existing models often struggle to generate images with the desired count of objects, visual text, or grounded emotions. We hope that our benchmark help ease future text-to-image generation research. The code and data are available at https://eslambakr.github.io/hrsbench.github.io

Agentic systems are evaluated on benchmarks where agents interact with environments to solve tasks. Most papers report a pass@1 score computed from a single run per task, assuming this gives a reliable performance estimate. We test this assumption by collecting 60,000 agentic trajectories on SWE-Bench-Verified, spanning three models and two scaffolds. We find substantial variance: single-run pass@1 estimates vary by 2.2 to 6.0 percentage points depending on which run is selected, with standard deviations exceeding 1.5 percentage points even at temperature 0. This variance has critical implications: reported improvements of 2--3 percentage points may reflect evaluation noise rather than genuine algorithmic progress. Through token-level analysis, we show that trajectories diverge early, often within the first few percent of tokens, and that these small differences cascade into different solution strategies. To enable reliable evaluation of agentic systems, we recommend three concrete practices: (1) estimate pass@1 from multiple independent runs per task, especially when measuring small improvements, (2) use statistical power analysis to determine the number of runs needed to detect expected effect sizes, and (3) consider metrics like pass@k (optimistic bound) and pass^k (pessimistic bound) with k>1 to better characterize the full performance envelope. While these practices increase evaluation cost, they are essential for distinguishing genuine scientific progress from statistical noise.

We present an automated, contrastive evaluation pipeline for auditing the behavioral impact of interventions on large language models. Given a base model M_1 and an intervention model M_2, our method compares their free-form, multi-token generations across aligned prompt contexts and produces human-readable, statistically validated natural-language hypotheses describing how the models differ, along with recurring themes that summarize patterns across validated hypotheses. We evaluate the approach in synthetic setting by injecting known behavioral changes and showing that the pipeline reliably recovers them. We then apply it to three real-world interventions, reasoning distillation, knowledge editing and unlearning, demonstrating that the method surfaces both intended and unexpected behavioral shifts, distinguishes large from subtle interventions, and does not hallucinate differences when effects are absent or misaligned with the prompt bank. Overall, the pipeline provides a statistically grounded and interpretable tool for post-hoc auditing of intervention-induced changes in model behavior.

Practitioners make three prompt-design decisions with almost no controlled evidence behind them: how to format instructions and context (markdown, plain text, prose, or tabular), how many simultaneous instructions a system prompt can carry before compliance degrades, and how much context a model can hold before recall and honesty degrade. We report two controlled experiments crossing all three factors on one held, contamination-free synthetic corpus (the "Book of Veyra," 8,780 uniquely-named entities, deterministically regenerable from a fixed seed), evaluated across five models. Experiment 1 (960 calls/model) measures instruction-following decay as rule count N grows from 10 to 160, crossed with four formats and system-prompt vs. user-turn placement. Perfect-response rate collapses to zero by N=80 for every model, format, and placement. Placement produces effects at least as large as format at N=160 in most models, but the direction is model-specific. No model shows a reliable markdown advantage; one 35B model favors plain text instead. Experiment 2 (5,520 calls/model) measures recall accuracy, false-premise sycophancy, and absent-fact fabrication across a 2k-to-512k-token context ladder in the same four formats. Recall stays near ceiling through 64-128k tokens, then degrades sharply and format-dependently: one model's accuracy spread reaches 48 points at 128k tokens. Fabrication never occurs (0/5,760 probes), and sycophancy stays negligible (<=8.3%). What rises sharply near each model's context ceiling is outright refusal to answer (0% to 79-90%), distinct from sycophancy or fabrication. Neither pre-registered format ordering holds, and token overhead (+22% to +37% over plain text) further changes which format is preferable where accuracy spread is genuine. We release the full harness, corpus generator, and raw results (VeyraBench): https://github.com/iNetanel/veyrabench

Observational studies can yield clinically actionable evidence at scale, but executing them on real-world databases is open-ended and requires coherent decisions across cohort construction, analysis, and reporting. Prior evaluations of LLM agents emphasize isolated steps or single answers, missing the integrity and internal structure of the resulting evidence bundle. To address this gap, we introduce RWE-bench, a benchmark grounded in MIMIC-IV and derived from peer-reviewed observational studies. Each task provides the corresponding study protocol as the reference standard, requiring agents to execute experiments in a real database and iteratively generate tree-structured evidence bundles. We evaluate six LLMs (three open-source, three closed-source) under three agent scaffolds using both question-level correctness and end-to-end task metrics. Across 162 tasks, task success is low: the best agent reaches 39.9%, and the best open-source model reaches 30.4%. Agent scaffolds also matter substantially, causing over 30% variation in performance metrics. Furthermore, we implement an automated cohort evaluation method to rapidly localize errors and identify agent failure modes. Overall, the results highlight persistent limitations in agents' ability to produce end-to-end evidence bundles, and efficient validation remains an important direction for future work. Code and data are available at https://github.com/somewordstoolate/RWE-bench.

Large language models (LLMs) face a dual challenge in creative capability evaluation: existing benchmarks (e.g., Story Cloze Test, HellaSwag) measure models' discriminative ability over narrative continuation using multiple-choice recognition paradigms, rather than directly measuring creative generation capability; rubric-based scoring and LLM-as-Judge methods rely on subjective dimension assessment or natural language model outputs, and cannot provide objective, automated scoring mechanisms. This paper proposes QUIET (Quality Understanding via Interlocked Evaluation Testing), a diagnostic benchmark for LLM creative capability based on multi-blank cascaded story cloze. QUIET sets N blanks (10-20) in a story with complete structure, with each blank accompanied by an explicit content constraint, and cascade dependency relationships between blanks -- the content filled into earlier blanks constrains the feasible solution space for later blanks. The evaluated model (or human participants) fills all blanks in open-ended generation mode; the results are scored by an information-theoretic automated scoring protocol without human grading. The scoring protocol directly operationalizes the "calibrated surprise" theoretical framework (Zou & Xu, 2026a). For each blank k, a composite score is computed: score = satisfy * (1 + lambda * surprise), where lambda = 1.0. Here, "satisfy" measures how well the blank filling satisfies the content constraint (objective logical reasoning judgment, not subjective aesthetic scoring), and "surprise" measures the degree of surprise given that the constraint is satisfied. Creative answers that do not satisfy the constraint score zero; answers that satisfy the constraint but are mediocre score low; answers that satisfy the constraint and are surprising score high.

Evaluations should do more than measure a models current performance. They should tell us what to fix for the next model iteration and provide a way to generate targeted post training data. Most evaluation pipelines identify weak examples, topics, or categories, but they leave the underlying capability failure implicit: they say where a model fails, not why. We introduce CRAFT, a method that converts any rubric based evaluation dataset into a model specific diagnosis of weak capabilities. CRAFT treats each grading criterion as a capability probe: it extracts a capability description from every prompt rubric pair, clusters these descriptions into a hierarchical capability tree, scores the target model at every node, and selects low performing nodes dynamically across tree levels, at the granularity where each failure is clearest. The selected weak capabilities then direct the generation of targeted supervised finetuning data. Holding the data generation, finetuning, and evaluation setup fixed, we compare CRAFT against prompt level EvalTree clustering and untargeted random generation on four open source models, two professional domains (finance and legal), and 13 held out benchmarks disjoint from the diagnostic data. CRAFT achieves the strongest finance domain average for all four models under repeated temperature decoding; on legal domain, it is strongest for three of four models and remains within the decoding variance bands of the best baseline on the fourth. Diagnosing weaknesses at the level of rubric criteria, rather than prompts or categories, thus yields both a sharper picture of what a model cannot do and measurably better models after finetuning on that diagnosis.

Evaluation harnesses are software systems that orchestrate model evaluation by managing model invocation, data loading, metric computation, and result reporting. Despite their critical role in machine learning infrastructure, their operational challenges and engineering concerns have received limited attention so far. We present an empirical study of 57 evaluation harnesses, deriving a five-stage harness model and classifying 16,560 issues by workflow stage and root cause. Most harness operational challenges concentrate in the Specification stage (41.4% of issues), where harnesses integrate external models, datasets, and scoring judges. The three most frequent root causes of operational challenges are unimplemented features (24.3%), documentation gaps (20.3%), and missing input validation (17.2%), which together account for 61.7% of classified issues, spanning both defects in existing functionality and capability gaps that block intended workflows. Root causes also vary by workflow stage: environment incompatibility and external dependency breakage account for 36.2% of provisioning issues, whereas algorithmic error (25.9%) and validation gap (22.5%) dominate assessment issues. Together, these contributions establish an empirical foundation for treating evaluation engineering as a distinct software engineering concern.

This position paper argues that computer science conferences should require tamper-evident, nonrepudiable attestations of experimental results. We name the underlying problem experiment nonrepudiation: a compliant protocol must bind the numbers in a paper to an actual executed computation in a way the author cannot later alter or deny. The current system relies on self-reported checklists, optional code sharing, and author-controlled logging. None of these mechanisms answer the question a reviewer cannot check: did the code the paper describes produce the numbers the paper reports? We define the problem formally, state the security properties any compliant protocol must satisfy, and describe a threat model that includes attacks current approaches do not prevent. To show that the problem is solvable, we built K-Veritas, a reference implementation in Go that produces signed reports without accessing training data. K-Veritas is a testbed, not a finished answer. We call on conferences and the community to treat nonrepudiation as a first-class requirement and to help build an open, independent standard for it.

We present HealthBench, an open-source benchmark measuring the performance and safety of large language models in healthcare. HealthBench consists of 5,000 multi-turn conversations between a model and an individual user or healthcare professional. Responses are evaluated using conversation-specific rubrics created by 262 physicians. Unlike previous multiple-choice or short-answer benchmarks, HealthBench enables realistic, open-ended evaluation through 48,562 unique rubric criteria spanning several health contexts (e.g., emergencies, transforming clinical data, global health) and behavioral dimensions (e.g., accuracy, instruction following, communication). HealthBench performance over the last two years reflects steady initial progress (compare GPT-3.5 Turbo's 16% to GPT-4o's 32%) and more rapid recent improvements (o3 scores 60%). Smaller models have especially improved: GPT-4.1 nano outperforms GPT-4o and is 25 times cheaper. We additionally release two HealthBench variations: HealthBench Consensus, which includes 34 particularly important dimensions of model behavior validated via physician consensus, and HealthBench Hard, where the current top score is 32%. We hope that HealthBench grounds progress towards model development and applications that benefit human health.

Many deployments must compare candidate language models for safety before a labeled benchmark exists for the relevant language, sector, or regulatory regime. We formalize this setting as benchmarkless comparative safety scoring and specify the contract under which a scenario-based audit can be interpreted as deployment evidence. Scores are valid only under a fixed scenario pack, rubric, auditor, judge, sampling configuration, and rerun budget. Because no labels are available, we replace ground-truth agreement with an instrumental-validity chain: responsiveness to a controlled safe-versus-abliterated contrast, dominance of target-driven variance over auditor and judge artifacts, and stability across reruns. We instantiate the chain in SimpleAudit, a local-first scoring instrument, and validate it on a Norwegian safety pack. Safe and abliterated targets separate with AUROC values between 0.89 and 1.00, target identity is the dominant variance component (η^2 approx 0.52), and severity profiles stabilize by ten reruns. Applying the same chain to Petri shows that it admits both tools. The substantial differences arise upstream of the chain, in claim-contract enforcement and deployment fit. A Norwegian public-sector procurement case comparing Borealis and Gemma 3 demonstrates the resulting evidence in practice: the safer model depends on scenario category and risk measure. Consequently, scores, matched deltas, critical rates, uncertainty, and the auditor and judge used must be reported together rather than collapsed into a single ranking.

Human evaluation is the gold standard for multilingual NLP, but is often skipped in practice and substituted with automatic metrics, because it is notoriously complex and slow to set up with existing tools with substantial engineering and operational overhead. We introduce Pearmut, a lightweight yet feature-rich platform that makes end-to-end human evaluation as easy to run as automatic evaluation. Pearmut removes common entry barriers and provides support for evaluating multilingual tasks, with a particular focus on machine translation. The platform implements standard evaluation protocols, including DA, ESA, or MQM, but is also extensible to allow prototyping new protocols. It features document-level context, absolute and contrastive evaluation, attention checks, ESAAI pre-annotations and both static and active learning-based assignment strategies. Pearmut enables reliable human evaluation to become a practical, routine component of model development and diagnosis rather than an occasional effort.

Text-to-image (T2I) models have rapidly advanced, enabling the generation of high-quality images from text prompts across various domains. However, these models present notable safety concerns, including the risk of generating harmful, biased, or private content. Current research on assessing T2I safety remains in its early stages. While some efforts have been made to evaluate models on specific safety dimensions, many critical risks remain unexplored. To address this gap, we introduce T2ISafety, a safety benchmark that evaluates T2I models across three key domains: toxicity, fairness, and bias. We build a detailed hierarchy of 12 tasks and 44 categories based on these three domains, and meticulously collect 70K corresponding prompts. Based on this taxonomy and prompt set, we build a large-scale T2I dataset with 68K manually annotated images and train an evaluator capable of detecting critical risks that previous work has failed to identify, including risks that even ultra-large proprietary models like GPTs cannot correctly detect. We evaluate 12 prominent diffusion models on T2ISafety and reveal several concerns including persistent issues with racial fairness, a tendency to generate toxic content, and significant variation in privacy protection across the models, even with defense methods like concept erasing. Data and evaluator are released under https://github.com/adwardlee/t2i_safety.

Due to the cumbersome nature of human evaluation and limitations of code-based evaluation, Large Language Models (LLMs) are increasingly being used to assist humans in evaluating LLM outputs. Yet LLM-generated evaluators simply inherit all the problems of the LLMs they evaluate, requiring further human validation. We present a mixed-initiative approach to ``validate the validators'' -- aligning LLM-generated evaluation functions (be it prompts or code) with human requirements. Our interface, EvalGen, provides automated assistance to users in generating evaluation criteria and implementing assertions. While generating candidate implementations (Python functions, LLM grader prompts), EvalGen asks humans to grade a subset of LLM outputs; this feedback is used to select implementations that better align with user grades. A qualitative study finds overall support for EvalGen but underscores the subjectivity and iterative process of alignment. In particular, we identify a phenomenon we dub criteria drift: users need criteria to grade outputs, but grading outputs helps users define criteria. What is more, some criteria appears dependent on the specific LLM outputs observed (rather than independent criteria that can be defined a priori), raising serious questions for approaches that assume the independence of evaluation from observation of model outputs. We present our interface and implementation details, a comparison of our algorithm with a baseline approach, and implications for the design of future LLM evaluation assistants.

Prompt design is a primary control interface for large language models (LLMs), yet standard evaluations largely reduce performance to answer correctness, obscuring why a prompt succeeds or fails and providing little actionable guidance. We propose PEEM (Prompt Engineering Evaluation Metrics), a unified framework for joint and interpretable evaluation of both prompts and responses. PEEM defines a structured rubric with 9 axes: 3 prompt criteria (clarity/structure, linguistic quality, fairness) and 6 response criteria (accuracy, coherence, relevance, objectivity, clarity, conciseness), and uses an LLM-based evaluator to output (i) scalar scores on a 1-5 Likert scale and (ii) criterion-specific natural-language rationales grounded in the rubric. Across 7 benchmarks and 5 task models, PEEM's accuracy axis strongly aligns with conventional accuracy while preserving model rankings (aggregate Spearman rho about 0.97, Pearson r about 0.94, p < 0.001). A multi-evaluator study with four models shows consistent relative judgments (pairwise rho = 0.68-0.85), supporting evaluator-agnostic deployment. Beyond alignment, PEEM captures complementary linguistic failure modes and remains informative under prompt perturbations: prompt-quality trends track downstream accuracy under iterative rewrites, semantic adversarial manipulations induce clear score degradation, and meaning-preserving paraphrases yield high stability (robustness rate about 76.7-80.6%). Finally, using only PEEM scores and rationales as feedback, a zero-shot prompt rewriting loop improves downstream accuracy by up to 11.7 points, outperforming supervised and RL-based prompt-optimization baselines. Overall, PEEM provides a reproducible, criterion-driven protocol that links prompt formulation to response behavior and enables systematic diagnosis and optimization of LLM interactions.

Medical large language models (LLMs) research often makes bold claims, from encoding clinical knowledge to reasoning like a physician. These claims are usually backed by evaluation on competitive benchmarks; a tradition inherited from mainstream machine learning. But how do we separate real progress from a leaderboard flex? Medical LLM benchmarks, much like those in other fields, are arbitrarily constructed using medical licensing exam questions. For these benchmarks to truly measure progress, they must accurately capture the real-world tasks they aim to represent. In this position paper, we argue that medical LLM benchmarks should (and indeed can) be empirically evaluated for their construct validity. In the psychological testing literature, "construct validity" refers to the ability of a test to measure an underlying "construct", that is the actual conceptual target of evaluation. By drawing an analogy between LLM benchmarks and psychological tests, we explain how frameworks from this field can provide empirical foundations for validating benchmarks. To put these ideas into practice, we use real-world clinical data in proof-of-concept experiments to evaluate popular medical LLM benchmarks and report significant gaps in their construct validity. Finally, we outline a vision for a new ecosystem of medical LLM evaluation centered around the creation of valid benchmarks.

Generative AI systems across modalities, ranging from text (including code), image, audio, and video, have broad social impacts, but there is no official standard for means of evaluating those impacts or for which impacts should be evaluated. In this paper, we present a guide that moves toward a standard approach in evaluating a base generative AI system for any modality in two overarching categories: what can be evaluated in a base system independent of context and what can be evaluated in a societal context. Importantly, this refers to base systems that have no predetermined application or deployment context, including a model itself, as well as system components, such as training data. Our framework for a base system defines seven categories of social impact: bias, stereotypes, and representational harms; cultural values and sensitive content; disparate performance; privacy and data protection; financial costs; environmental costs; and data and content moderation labor costs. Suggested methods for evaluation apply to listed generative modalities and analyses of the limitations of existing evaluations serve as a starting point for necessary investment in future evaluations. We offer five overarching categories for what can be evaluated in a broader societal context, each with its own subcategories: trustworthiness and autonomy; inequality, marginalization, and violence; concentration of authority; labor and creativity; and ecosystem and environment. Each subcategory includes recommendations for mitigating harm.

Safety and responsibility evaluations of advanced AI models are a critical but developing field of research and practice. In the development of Google DeepMind's advanced AI models, we innovated on and applied a broad set of approaches to safety evaluation. In this report, we summarise and share elements of our evolving approach as well as lessons learned for a broad audience. Key lessons learned include: First, theoretical underpinnings and frameworks are invaluable to organise the breadth of risk domains, modalities, forms, metrics, and goals. Second, theory and practice of safety evaluation development each benefit from collaboration to clarify goals, methods and challenges, and facilitate the transfer of insights between different stakeholders and disciplines. Third, similar key methods, lessons, and institutions apply across the range of concerns in responsibility and safety - including established and emerging harms. For this reason it is important that a wide range of actors working on safety evaluation and safety research communities work together to develop, refine and implement novel evaluation approaches and best practices, rather than operating in silos. The report concludes with outlining the clear need to rapidly advance the science of evaluations, to integrate new evaluations into the development and governance of AI, to establish scientifically-grounded norms and standards, and to promote a robust evaluation ecosystem.

Deep research systems are widely used for multi-step web research, analysis, and cross-source synthesis, yet their evaluation remains challenging. Existing benchmarks often require annotation-intensive task construction, rely on static evaluation dimensions, or fail to reliably verify facts when citations are missing. To bridge these gaps, we introduce DeepResearchEval, an automated framework for deep research task construction and agentic evaluation. For task construction, we propose a persona-driven pipeline generating realistic, complex research tasks anchored in diverse user profiles, applying a two-stage filter Task Qualification and Search Necessity to retain only tasks requiring multi-source evidence integration and external retrieval. For evaluation, we propose an agentic pipeline with two components: an Adaptive Point-wise Quality Evaluation that dynamically derives task-specific evaluation dimensions, criteria, and weights conditioned on each generated task, and an Active Fact-Checking that autonomously extracts and verifies report statements via web search, even when citations are missing.

The recent rapid advancement of Text-to-Video (T2V) generation technologies are engaging the trained models with more world model ability, making the existing benchmarks increasingly insufficient to evaluate state-of-the-art T2V models. First, current evaluation dimensions, such as per-frame aesthetic quality and temporal consistency, are no longer able to differentiate state-of-the-art T2V models. Second, event-level temporal causality-an essential property that differentiates videos from other modalities-remains largely unexplored. Third, existing benchmarks lack a systematic assessment of world knowledge, which are essential capabilities for building world models. To address these issues, we introduce VideoVerse, a comprehensive benchmark focusing on evaluating whether the current T2V model could understand complex temporal causality and world knowledge to synthesize videos. We collect representative videos across diverse domains and extract their event-level descriptions with inherent temporal causality, which are then rewritten into text-to-video prompts by independent annotators. For each prompt, we design ten evaluation dimensions covering dynamic and static properties, resulting in 300 prompts, 815 events, and 793 evaluation questions. Consequently, a human preference-aligned QA-based evaluation pipeline is developed by using modern vision-language models to systematically benchmark leading open- and closed-source T2V systems, revealing the current gap between T2V models and desired world modeling abilities.

The Model Context Protocol (MCP) is rapidly becoming the standard interface for Large Language Models (LLMs) to discover and invoke external tools. However, existing evaluations often fail to capture the complexity of real-world scenarios, relying on restricted toolsets, simplistic workflows, or subjective LLM-as-a-judge metrics. We introduce MCP-Atlas, a large-scale benchmark for evaluating tool-use competency, comprising 36 real MCP servers and 220 tools. It includes 1,000 tasks designed to assess tool-use competency in realistic, multi-step workflows. Tasks use natural language prompts that avoid naming specific tools or servers, requiring agents to identify and orchestrate 3-6 tool calls across multiple servers. We score tasks using a claims-based rubric that awards partial credit based on the factual claims satisfied in the model's final answer, complemented by internal diagnostics on tool discovery, parameterization, syntax, error recovery, and efficiency. Evaluation results on frontier models reveal that top models achieve pass rates exceeding 50%, with primary failures arising from inadequate tool usage and task understanding. We release the task schema, containerized harness, and a 500-task public subset of the benchmark dataset to facilitate reproducible comparisons and advance the development of robust, tool-augmented agents.

An isolated final user message is often treated as the query in evaluations of AI systems. In a conversation, however, the actionable request may be distributed across preceding turns. We directly test whether that omitted within-conversation context changes answers. For each of 180 English multi-turn conversations sampled from a governed commercial corpus and the public PRISM dataset, we hold the final user message and requested answer model constant while generating three answers: one from the full role-labelled conversation, one from the final message alone, and one from the final message plus a prefix-only reconstruction capped at 160 words. A separately requested judge model evaluates answers under randomized labels. The prespecified primary endpoint is a material difference that could change what the user does, rather than a difference in style or detail. After inverse-probability weighting to the eligible cohorts, the full-conversation and isolated-final answers differ materially in 44.7% of cases (95% bootstrap CI 33.8% to 56.1%). Full-conversation answers score 0.49 points higher on a 0 to 4 request-satisfaction scale (0.32 to 0.67). Adding the compressed prefix reduces the material-difference rate to 30.8% (20.2% to 42.1%), a 13.9-point reduction (4.9% to 24.1%), and reduces the mean satisfaction gap to 0.01 points (-0.12 to 0.13). Yet compression is not equivalent to the complete dialogue context: almost one third of answers remain materially different. An order-swapped repeat on 48 cases yields 91.7% agreement and kappa = 0.83 for the primary decision. The study concerns preceding turns in the same conversation and does not test persistent memory across separate conversations.

As research in large language models (LLMs) continues to accelerate, LLM-based evaluation has emerged as a scalable and cost-effective alternative to human evaluations for comparing the ever increasing list of models. This paper investigates the efficacy of these "LLM evaluators", particularly in using them to assess instruction following, a metric that gauges how closely generated text adheres to the given instruction. We introduce a challenging meta-evaluation benchmark, LLMBar, designed to test the ability of an LLM evaluator in discerning instruction-following outputs. The authors manually curated 419 pairs of outputs, one adhering to instructions while the other diverging, yet may possess deceptive qualities that mislead an LLM evaluator, e.g., a more engaging tone. Contrary to existing meta-evaluation, we discover that different evaluators (i.e., combinations of LLMs and prompts) exhibit distinct performance on LLMBar and even the highest-scoring ones have substantial room for improvement. We also present a novel suite of prompting strategies that further close the gap between LLM and human evaluators. With LLMBar, we hope to offer more insight into LLM evaluators and foster future research in developing better instruction-following models.

As LLM-based agents are deployed in complex, multi-step workflows, a critical evaluation gap has emerged: most existing benchmarks judge only final outcomes, unable to distinguish reliable reasoning from lucky success or attribute failures to specific process deficiencies, hindering attribution in long-horizon tasks. In this work, we present ClawTrack, a dual-assessment benchmark that simultaneously measures what an agent achieves (Task Score) and how it achieves it (Process Score). ClawTrack comprises 320 tasks across 8 domains with 25+ deterministic mock services. A Process Grader scores each reasoning turn along four dimensions (goal alignment, efficiency, information utilization, and result verification), anchored by 12,541 task-specific rubric items. Evaluating 21 models over 16,000+ trials, we find that: (1) process scores effectively attribute success and failure to specific reasoning dimensions, filtering lucky passes invisible to outcome-only evaluation; (2) the four dimensions are complementary, with result verification as the systematic bottleneck; (3) the framework is robust to evaluator choice across different judge LLMs; and (4) process-based trajectory filtering yields consistent post-training improvements across model scales.

This paper presents synthetic Preference Optimization (PO) datasets generated using multi-agent workflows and evaluates the effectiveness and potential of these workflows in the dataset generation process. PO dataset generation requires two modules: (1) response evaluation, and (2) response generation. In the response evaluation module, the responses from Large Language Models (LLMs) are evaluated and ranked - a task typically carried out by human annotators that we automate using LLMs. We assess the response evaluation module in a 2 step process. In step 1, we assess LLMs as evaluators using three distinct prompting strategies. In step 2, we apply the winning prompting strategy to compare the performance of LLM-as-a-Judge, LLMs-as-a-Jury, and LLM Debate. In each step, we use inter-rater agreement using Cohen's Kappa between human annotators and LLMs. For the response generation module, we compare different configurations for the LLM Feedback Loop using the identified LLM evaluator configuration. We use the win rate (the fraction of times a generation framework is selected as the best by an LLM evaluator) to determine the best multi-agent configuration for generation. After identifying the best configurations for both modules, we use models from the GPT, Gemma, and Llama families to generate our PO datasets using the above pipeline. We generate two types of PO datasets, one to improve the generation capabilities of individual LLM and the other to improve the multi-agent workflow. Our evaluation shows that GPT-4o-as-a-Judge is more consistent across datasets when the candidate responses do not include responses from the GPT family. Additionally, we find that the LLM Feedback Loop, with Llama as the generator and Gemma as the reviewer, achieves a notable 71.8% and 73.8% win rate over single-agent Llama and Gemma, respectively.

AI evaluations are widely used for testing and understanding progress. However, the diverse evaluators bring with them inconsistencies that challenge analysis and comparison. First, results are saved in incompatible formats, scattered across leaderboards, papers, blog posts, evaluation harness logs, and custom repositories. Second, results are created by different evaluation frameworks, which produce divergent scores for nominally identical evaluations and record metadata inconsistently, hindering comparison, cross-community evaluation science, cost reduction, and reuse. We introduce Every Eval Ever, the first shared schema and community-crowdsourced repository for AI evaluation results. The schema standardizes how evaluations are represented in a unified, single JSON document. It is source-agnostic by design, ingesting results from evaluation harnesses and papers alike, and optionally stores per-instance outputs for fine-grained analysis. We contribute: (i) a community-governed metadata schema with a companion instance-level schema, the first standardization effort of its kind; (ii) automatic converters from popular formats, evaluation harnesses, and leaderboards to the unified schema; and (iii) a crowdsourced community database hosted on Hugging Face, currently spanning to date 22,235 models, 2,273 unique benchmarks, and 31 evaluation formats.

Given the widespread adoption and usage of Large Language Models (LLMs), it is crucial to have flexible and interpretable evaluations of their instruction-following ability. Preference judgments between model outputs have become the de facto evaluation standard, despite distilling complex, multi-faceted preferences into a single ranking. Furthermore, as human annotation is slow and costly, LLMs are increasingly used to make these judgments, at the expense of reliability and interpretability. In this work, we propose TICK (Targeted Instruct-evaluation with ChecKlists), a fully automated, interpretable evaluation protocol that structures evaluations with LLM-generated, instruction-specific checklists. We first show that, given an instruction, LLMs can reliably produce high-quality, tailored evaluation checklists that decompose the instruction into a series of YES/NO questions. Each question asks whether a candidate response meets a specific requirement of the instruction. We demonstrate that using TICK leads to a significant increase (46.4% to 52.2%) in the frequency of exact agreements between LLM judgements and human preferences, as compared to having an LLM directly score an output. We then show that STICK (Self-TICK) can be used to improve generation quality across multiple benchmarks via self-refinement and Best-of-N selection. STICK self-refinement on LiveBench reasoning tasks leads to an absolute gain of +7.8%, whilst Best-of-N selection with STICK attains +6.3% absolute improvement on the real-world instruction dataset, WildBench. In light of this, structured, multi-faceted self-improvement is shown to be a promising way to further advance LLM capabilities. Finally, by providing LLM-generated checklists to human evaluators tasked with directly scoring LLM responses to WildBench instructions, we notably increase inter-annotator agreement (0.194 to 0.256).

Despite the utility of Large Language Models (LLMs) across a wide range of tasks and scenarios, developing a method for reliably evaluating LLMs across varied contexts continues to be challenging. Modern evaluation approaches often use LLMs to assess responses generated by LLMs. However, the meta-evaluation conducted to assess the effectiveness of these LLMs as evaluators is typically constrained by the coverage of existing benchmarks or requires extensive human annotation. This underscores the urgency of methods for scalable meta-evaluation that can effectively, reliably, and efficiently evaluate the performance of LLMs as evaluators across diverse tasks and scenarios, particularly in potentially new, user-defined scenarios. To fill this gap, we propose ScaleEval, an agent-debate-assisted meta-evaluation framework that leverages the capabilities of multiple communicative LLM agents. This framework supports multi-round discussions to assist human annotators in discerning the most capable LLMs as evaluators, which significantly eases their workload in cases that used to require large-scale annotations during meta-evaluation. We release the code for our framework, which is publicly available at: https://github.com/GAIR-NLP/scaleeval.

If AI models can detect when they are being evaluated, the effectiveness of evaluations might be compromised. For example, models could have systematically different behavior during evaluations, leading to less reliable benchmarks for deployment and governance decisions. We investigate whether frontier language models can accurately classify transcripts based on whether they originate from evaluations or real-world deployment, a capability we call evaluation awareness. To achieve this, we construct a diverse benchmark of 1,000 prompts and transcripts from 61 distinct datasets. These span public benchmarks (e.g., MMLU, SWEBench), real-world deployment interactions, and agent trajectories from scaffolding frameworks (e.g., web-browsing agents). Frontier models clearly demonstrate above-random evaluation awareness (Gemini-2.5-Pro reaches an AUC of 0.83), but do not yet surpass our simple human baseline (AUC of 0.92). Furthermore, both AI models and humans are better at identifying evaluations in agentic settings compared to chat settings. Additionally, we test whether models can identify the purpose of the evaluation. Under multiple-choice and open-ended questioning, AI models far outperform random chance in identifying what an evaluation is testing for. Our results indicate that frontier models already exhibit a substantial, though not yet superhuman, level of evaluation-awareness. We recommend tracking this capability in future models.

In medical education, physicians convert academic knowledge into clinical expertise through residency: years of training across thousands of encounters, with diverse sources of feedback and progressively greater autonomy. Much of clinical reasoning relies on the patient encounter, a dialogue in which a clinician elicits history, refines diagnostic hypotheses, and decides management under uncertainty. While large language models (LLMs) excel on static medical benchmarks, methods to optimize the full sequence of clinical decisions remain underdeveloped. We present ResidencyRL, a reinforcement learning (RL) method for training clinical artificial intelligence (AI) agents through simulated multi-turn clinical encounters (up to 60 dialogue turns and 8 tool calls per trajectory). ResidencyRL pairs the policy agent with LLM simulators capable of complex, adversarial behaviors, training against a structured reward aligned to diagnostic accuracy, management quality, communication, documentation, and safety. On held-out evaluations, the ResidencyRL agent improves diagnostic accuracy by 7.0% under adversarial conditions (88.0% vs. 81.0%) and reduces missed red flag rates by 31%, demonstrating rigorous mitigation of premature closure. Blinded expert clinicians validated these gains, preferring the trained agent in 87.6% of side-by-side comparisons. The procedural competencies transfer to unseen benchmarks: the agent outperforms the base model across all six clinical axes of the AMIE multi-visit benchmark, and shows consistent directional improvements on AgentClinic and CRAFT-MD. Our findings demonstrate that sequential clinical decision-making can be effectively learned through multi-turn RL in simulation, yielding robust, generalizable capabilities, paving the way towards clinical mastery. Prospective validation with real-world workflows remains necessary to establish clinical utility.

Scaling test-time computation, or affording a generator large language model (LLM) extra compute during inference, typically employs the help of external non-generative evaluators (i.e., reward models). Concurrently, LLM-judges, models trained to generate evaluations and critiques (explanations) in natural language, are becoming increasingly popular in automatic evaluation. Despite judge empirical successes, their effectiveness as evaluators in test-time scaling settings is largely unknown. In this paper, we introduce the Judge Evaluation for Test-Time Scaling (JETTS) benchmark, which evaluates judge performance in three domains (math reasoning, code generation, and instruction following) under three task settings: response reranking, step-level beam search, and critique-based response refinement. We evaluate 10 different judge models (7B-70B parameters) for 8 different base generator models (6.7B-72B parameters). Our benchmark shows that while judges are competitive with outcome reward models in reranking, they are consistently worse than process reward models in beam search procedures. Furthermore, though unique to LLM-judges, their natural language critiques are currently ineffective in guiding the generator towards better responses.