Daily Papers

Rubric-based reinforcement learning extends RL beyond tasks with exact answers or rule-based verifiers by scoring responses against instance-specific criteria. However, this makes reward computation expensive: training requires repeated rubric judging, often with proprietary APIs or local generative LLM judges with 7B parameters or more. We study whether smaller language models can serve as efficient and reliable rubric-based judges. To make this question measurable, we construct PointRubric and RaR-Science-Static, two pointwise rubric-based evaluation datasets with instance-specific criteria and itemwise satisfaction labels. We compare three ways of extracting criterion-level judgments from small models: Generative verdicts, Yes/No Logprob margins, and Probe judges. Across both datasets, the Qwen3-1.7B Probe judge achieves the strongest criterion-level agreement among these methods, outperforming Generative and Logprob judges. Used as a GRPO reward model, it trains a policy from 0.232 to 0.643 on RaR-Science rubric score, compared with 0.594 for an 8B Generative judge baseline, while the baseline requires 10.7times more reward-judge time. Task and domain transfer experiments further suggest that Probe judges preserve criterion-level reward structure across settings.

Claims that language models homogenise are usually measured against human judgements collected for the study, which makes the human side an artifact of the design: a crowdworker given the model's instruction is running the model's prompt. We measure convergence against a human reference nobody built for the purpose -- 2,523 reader mark sets across 120 web documents, produced by people highlighting for their own reasons on a platform where the overlay of others' marks is off by default. Agreement is the overlap between two size-matched sentence sets minus the overlap expected when each is resampled within its own depth-and-length bands. The null's calibration is demonstrated, not asserted: every pair involving a random baseline lands within 0.006 of zero. On the median document each party names 14 sentences of 70; two readers share 4.1 and two models 8.7. Across 18 model arms spanning 11 vendors, 3 countries and both weight regimes, the median of 153 model pairs is +0.093 against a human yardstick of +0.040, and 99 sit entirely above the human interval. Two frontier models from rival labs reach +0.203, twice what GPT-4o agrees with itself on a second call. The effect is not determinism, prompt wording, procedure, vendor or routing, and it is graded: the smallest models agree at the human level. No model agrees with readers detectably more than a reader does, and at equal depth and length no surface feature separates their choices. The multiples are procedure-dependent and the ordering is not: models are cut to their sharpest set while a reader's is a random draw from what they marked, and blunting the models alike halves the gap without closing it. Tested out of sample on four models released after this analysis, against predictions fixed beforehand, none clears the human interval. A population simulated from several models is not several populations.

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.

Objective. Clinical AI documentation systems require evaluation methodologies that are clinically valid, economically viable, and sensitive to iterative changes. Methods requiring expert review per scoring instance are too slow and expensive for safe, iterative deployment. We present a case-specific, clinician-authored rubric methodology for clinical AI evaluation and examine whether LLM-generated rubrics can approximate clinician agreement. Materials and Methods. Twenty clinicians authored 1,646 rubrics for 823 clinical cases (736 real-world, 87 synthetic) across primary care, psychiatry, oncology, and behavioral health. Each rubric was validated by confirming that an LLM-based scoring agent consistently scored clinician-preferred outputs higher than rejected ones. Seven versions of an EHR-embedded AI agent for clinicians were evaluated across all cases. Results. Clinician-authored rubrics discriminated effectively between high- and low-quality outputs (median score gap: 82.9%) with high scoring stability (median range: 0.00%). Median scores improved from 84% to 95%. In later experiments, clinician-LLM ranking agreement (tau: 0.42-0.46) matched or exceeded clinician-clinician agreement (tau: 0.38-0.43), attributable to both ceiling compression and LLM rubric improvement. Discussion. This convergence supports incorporating LLM rubrics alongside clinician-authored ones. At roughly 1,000 times lower cost, LLM rubrics enable substantially greater evaluation coverage, while continued clinical authorship grounds evaluation in expert judgment. Ceiling compression poses a methodological challenge for future inter-rater agreement studies. Conclusion. Case-specific rubrics offer a path for clinical AI evaluation that preserves expert judgment while enabling automation at three orders lower cost. Clinician-authored rubrics establish the baseline against which LLM rubrics are validated.

Sarcasm detection poses a fundamental challenge in computational semantics, requiring models to resolve disparities between literal and intended meaning. The challenge is amplified in low-resource languages where annotated datasets are scarce or nonexistent. We present Yor-Sarc, the first gold-standard dataset for sarcasm detection in Yorùbá, a tonal Niger-Congo language spoken by over 50 million people. The dataset comprises 436 instances annotated by three native speakers from diverse dialectal backgrounds using an annotation protocol specifically designed for Yorùbá sarcasm by taking culture into account. This protocol incorporates context-sensitive interpretation and community-informed guidelines and is accompanied by a comprehensive analysis of inter-annotator agreement to support replication in other African languages. Substantial to almost perfect agreement was achieved (Fleiss' κ= 0.7660; pairwise Cohen's κ= 0.6732--0.8743), with 83.3% unanimous consensus. One annotator pair achieved almost perfect agreement (κ= 0.8743; 93.8% raw agreement), exceeding a number of reported benchmarks for English sarcasm research works. The remaining 16.7% majority-agreement cases are preserved as soft labels for uncertainty-aware modelling. Yor-Sarchttps://github.com/toheebadura/yor-sarc is expected to facilitate research on semantic interpretation and culturally informed NLP for low-resource African languages.

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.

Qualitative research faces a critical reliability challenge: traditional inter-rater agreement methods require multiple human coders, are time-intensive, and often yield moderate consistency. We present a multi-perspective validation framework for LLM-based thematic analysis that combines ensemble validation with dual reliability metrics: Cohen's Kappa (κ) for inter-rater agreement and cosine similarity for semantic consistency. Our framework enables configurable analysis parameters (1-6 seeds, temperature 0.0-2.0), supports custom prompt structures with variable substitution, and provides consensus theme extraction across any JSON format. As proof-of-concept, we evaluate three leading LLMs (Gemini 2.5 Pro, GPT-4o, Claude 3.5 Sonnet) on a psychedelic art therapy interview transcript, conducting six independent runs per model. Results demonstrate Gemini achieves highest reliability (κ= 0.907, cosine=95.3%), followed by GPT-4o (κ= 0.853, cosine=92.6%) and Claude (κ= 0.842, cosine=92.1%). All three models achieve a high agreement (κ> 0.80), validating the multi-run ensemble approach. The framework successfully extracts consensus themes across runs, with Gemini identifying 6 consensus themes (50-83% consistency), GPT-4o identifying 5 themes, and Claude 4 themes. Our open-source implementation provides researchers with transparent reliability metrics, flexible configuration, and structure-agnostic consensus extraction, establishing methodological foundations for reliable AI-assisted qualitative research.

Large multimodal models (LMMs) are increasingly adopted as judges in multimodal evaluation systems due to their strong instruction following and consistency with human preferences. However, their ability to follow diverse, fine-grained evaluation criteria remains underexplored. We develop Multi-Crit, a benchmark for evaluating multimodal judges on their capacity to follow pluralistic criteria and produce reliable criterion-level judgments. Covering both open-ended generation and verifiable reasoning tasks, Multi-Crit is built through a rigorous data curation pipeline that gathers challenging response pairs with multi-criterion human annotations. It further introduces three novel metrics for systematically assessing pluralistic adherence, criterion-switching flexibility, and the ability to recognize criterion-level preference conflicts. Comprehensive analysis of 25 LMMs reveals that 1) proprietary models still struggle to maintain consistent adherence to pluralistic criteria--especially in open-ended evaluation; 2) open-source models lag further behind in flexibly following diverse criteria; and 3) critic fine-tuning with holistic judgment signals enhances visual grounding but fails to generalize to pluralistic criterion-level judgment. Additional analyses on reasoning fine-tuning, test-time scaling, and boundary consistency between open-source and proprietary models further probe the limits of current multimodal judges. As a pioneering study, Multi-Crit lays the foundation for building reliable and steerable multimodal AI evaluation.

Many tasks revolve around editing a document, whether code or text. We formulate the revision similarity problem to unify a wide range of machine learning evaluation problems whose goal is to assess a revision to an existing document. We observe that revisions usually change only a small portion of an existing document, so the existing document and its immediate revisions share a majority of their content. We formulate five adequacy criteria for revision similarity measures, designed to align them with human judgement. We show that popular pairwise measures, like BLEU, fail to meet these criteria, because their scores are dominated by the shared content. They report high similarity between two revisions when humans would assess them as quite different. This is a fundamental flaw we address. We propose a novel static measure, Excision Score (ES), which computes longest common subsequence (LCS) to remove content shared by an existing document with the ground truth and predicted revisions, before comparing only the remaining divergent regions. This is analogous to a surgeon creating a sterile field to focus on the work area. We use approximation to speed the standard cubic LCS computation to quadratic. In code-editing evaluation, where static measures are often used as a cheap proxy for passing tests, we demonstrate that ES surpasses existing measures. When aligned with test execution on HumanEvalFix, ES improves over its nearest competitor, SARI, by 12% Pearson correlation and by >21% over standard measures like BLEU. The key criterion is invariance to shared context; when we perturb HumanEvalFix with increased shared context, ES' improvement over SARI increases to 20% and >30% over standard measures. ES also handles other corner cases that other measures do not, such as correctly aligning moved code blocks, and appropriately rewarding matching insertions or deletions.

Large language models (LLMs) are increasingly used to generate scientific reviews, yet existing evaluations rarely examine whether different providers align with both conference decisions and human reviewing priorities within the same controlled setting. We compare reviews from OpenAI GPT-5.4, Google Gemini 3.1 Pro Preview, and Anthropic Claude Opus 4.6 with human reviews and final decisions for 300 topic-matched ICLR 2026 submissions, equally divided among oral, poster, and rejected papers. Each model reviewed every paper using identical instructions and rating scales after decision information was removed. Our study contributes a cross-provider analysis of three complementary dimensions: alignment with broad and fine-grained decision categories, differences in recommendation-scale usage, and thematic agreement in identified weaknesses. All three LLMs distinguished accepted from rejected papers, but none reproduced the oral versus poster distinction present in human ratings. Scoring patterns were provider-specific: Gemini assigned systematically higher ratings, while OpenAI and Claude were closer to humans for rejected and poster papers but more critical of oral papers. Human and LLM reviews also differed in emphasis, with LLMs more frequently identifying missing baseline comparisons and humans more often raising computational-efficiency concerns. These results show that broad decision alignment does not imply agreement with finer human judgments or reviewing priorities.

We present a principled approach to provide LLM-based evaluation with a rigorous guarantee of human agreement. We first propose that a reliable evaluation method should not uncritically rely on model preferences for pairwise evaluation, but rather assess the confidence of judge models and selectively decide when to trust its judgement. We then show that under this selective evaluation framework, human agreement can be provably guaranteed -- such that the model evaluation aligns with that of humans to a user-specified agreement level. As part of our framework, we also introduce Simulated Annotators, a novel confidence estimation method that significantly improves judge calibration and thus enables high coverage of evaluated instances. Finally, we propose Cascaded Selective Evaluation, where we use cheaper models as initial judges and escalate to stronger models only when necessary -- again, while still providing a provable guarantee of human agreement. Experimental results show that Cascaded Selective Evaluation guarantees strong alignment with humans, far beyond what LLM judges could achieve without selective evaluation. For example, on a subset of Chatbot Arena where GPT-4 almost never achieves 80% human agreement, our method, even while employing substantially cost-effective models such as Mistral-7B, guarantees over 80% human agreement with almost 80% test coverage.

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.

Rubric-based evaluation is widely used in LLM benchmarks and training pipelines for open-ended, less verifiable tasks. While prior work has demonstrated the effectiveness of rubrics using downstream signals such as reinforcement learning outcomes, there remains no principled way to diagnose how a rubric itself fails from such aggregated or downstream signals alone. To address this gap, we introduce RIFT: RubrIc Failure mode Taxonomy, a taxonomy for systematically characterizing failure modes in rubric composition and design. RIFT consists of eight failure modes organized into three high-level categories: Reliability Failures, Content Validity Failures, and Consequential Validity Failures. RIFT is developed using grounded theory by iteratively annotating rubrics drawn from five diverse data sources spanning general instruction following, code generation, creative writing, and expert-level deep research, until no new failure modes are identified. We evaluate the consistency of the taxonomy by measuring agreement among independent human annotators, observing fair agreement overall (87% pairwise agreement and 0.64 average Cohen's kappa). Finally, to support scalable diagnosis, we propose automated rubric quality metrics and show that they align with human failure-mode annotations, achieving up to 0.925 F1.

Large language models (LLMs) exhibit degraded performance under prompt compression, but the mechanisms remain poorly understood. We introduce the Compression-Decay Comprehension Test (CDCT), a benchmark that independently measures constraint compliance (CC) and semantic accuracy (SA) across compression levels. We evaluate 9 frontier LLMs across 8 concepts using 5 compression levels from extreme (c=0.0, ~2 words) to none (c=1.0, ~135 words). A three-judge LLM jury achieves almost perfect inter-rater agreement on CC (Fleiss' appa=0.90). We observe a universal U-curve pattern in constraint compliance (97.2% prevalence), with violations peaking at medium compression (c=0.5, ~27 words). Counterintuitively, models perform better at extreme compression than medium lengths. The dimensions are statistically orthogonal (r=0.193, p=0.084), with constraint effects 2.9x larger than semantic effects. Experimental validation via RLHF ablation confirms our constraint salience hypothesis: removing "helpfulness" signals improves CC by 598% on average (71/72 trials, p<0.001), with 79% achieving perfect compliance. This demonstrates that RLHF-trained helpfulness behaviors are the dominant cause of constraint violations at medium compression. Reasoning models outperform efficient models by 27.5% (Cohen's d=0.96). Our findings reveal a fundamental tension between RLHF alignment and instruction-following, providing actionable guidelines for improving deployed systems.

The LLM-as-a-Judge paradigm promises scalable rubric-based evaluation, yet aligning frozen black-box models with human standards remains a challenge due to inherent generation stochasticity. We reframe judge alignment as a criteria transfer problem and isolate three recurrent failure modes: rubric instability caused by prompt sensitivity, unverifiable reasoning that lacks auditable evidence, and scale misalignment with human grading boundaries. To address these issues, we introduce RULERS (Rubric Unification, Locking, and Evidence-anchored Robust Scoring), a compiler-executor framework that transforms natural language rubrics into executable specifications. RULERS operates by compiling criteria into versioned immutable bundles, enforcing structured decoding with deterministic evidence verification, and applying lightweight Wasserstein-based post-hoc calibration, all without updating model parameters. Extensive experiments on essay and summarization benchmarks demonstrate that RULERS significantly outperforms representative baselines in human agreement, maintains strong stability against adversarial rubric perturbations, and enables smaller models to rival larger proprietary judges. Overall, our results suggest that reliable LLM judging requires executable rubrics, verifiable evidence, and calibrated scales rather than prompt phrasing alone. Code is available at https://github.com/LabRAI/Rulers.git.

The rapid adoption of Large Language Models (LLMs) has spurred interest in automated peer review; however, progress is currently stifled by benchmarks that treat reviewing primarily as a rating prediction task. We argue that the utility of a review lies in its textual justification--its arguments, questions, and critique--rather than a scalar score. To address this, we introduce Beyond Rating, a holistic evaluation framework that assesses AI reviewers across five dimensions: Content Faithfulness, Argumentative Alignment, Focus Consistency, Question Constructiveness, and AI-Likelihood. Notably, we propose a Max-Recall strategy to accommodate valid expert disagreement and introduce a curated dataset of paper with high-confidence reviews, rigorously filtered to remove procedural noise. Extensive experiments demonstrate that while traditional n-gram metrics fail to reflect human preferences, our proposed text-centric metrics--particularly the recall of weakness arguments--correlate strongly with rating accuracy. These findings establish that aligning AI critique focus with human experts is a prerequisite for reliable automated scoring, offering a robust standard for future research.

Ambient AI "scribe" systems promise to reduce clinical documentation burden, but automatic speech recognition (ASR) errors can remain unnoticed without careful review, and high-quality human reference transcripts are often unavailable for calibrating uncertainty. We investigate whether cross-model disagreement among heterogeneous ASR systems can act as a reference-free uncertainty signal to prioritize human verification in medical transcription workflows. Using 50 publicly available medical education audio clips (8 h 14 min), we transcribed each clip with eight ASR systems spanning commercial APIs and open-source engines. We aligned multi-model outputs, built consensus pseudo-references, and quantified token-level agreement using a majority-strength metric; we further characterized disagreements by type (content vs. punctuation/formatting) and assessed per-model agreement via leave-one-model-out (jackknife) consensus scoring. Inter-model reliability was low (ICC[2,1] = 0.131), indicating heterogeneous failure modes across systems. Across 76,398 evaluated token positions, 72.1% showed near-unanimous agreement (7-8 models), while 2.5% fell into high-risk bands (0-3 models), with high-risk mass varying from 0.7% to 11.4% across accent groups. Low-agreement regions were enriched for content disagreements, with the content fraction increasing from 53.9% to 73.9% across quintiles of high-risk mass. These results suggest that cross-model disagreement provides a sparse, localizable signal that can surface potentially unreliable transcript spans without human-verified references, enabling targeted review; clinical accuracy of flagged regions remains to be established.

Portuguese Information Retrieval (IR) lacks public datasets, and relevance assessment for specialized collections remains costly. While Large Language Models (LLMs) increasingly support relevance assessment, their reliability in non-English specialized domains remains unclear. We introduce NormasTCU (https://huggingface.co/datasets/LeandroRibeiro/NormasTCU), a Brazilian Portuguese IR dataset with 14,469 legal documents, 46 queries, and 3,048 human judgments over 812 query-document pairs. Using NormasTCU, we evaluated LLM-as-a-judge for relevance assessment by prompting three models with two prompt techniques to grade these pairs. We then compared the rankings of 15 IR systems derived from LLM-generated and human reference qrels. LLMs consistently showed a positive scoring bias (mean absolute error: 0.46--0.66 on a 0-2 scale). Furthermore, pair-level agreement with human judgments achieved only fair to moderate levels, with Cohen's kappa ranging from 0.32 to 0.53. Despite this bias, LLM-generated judgments often yielded highly similar system rankings for nDCG@10 and MRR (observed Kendall's tau greater than or equal 0.90, although the bootstrap confidence intervals did not always remain above this threshold), but were less reliable for P@10 and R@10. Notably, LLM-based rankings were sometimes more strongly correlated with the reference ranking than individual human annotations were. As a practical implication, our results suggest that LLMs could effectively support scalable relevance assessment in specialized Portuguese corpora when evaluated using nDCG or MRR (rank-aware metrics), but they should be avoided when relying on precision or recall.

Pairwise human preference prediction is central to evaluating code-generation systems, where quality often depends on task-specific trade-offs beyond functional correctness. While rubric-based LLM judges improve interpretability by decomposing evaluation into explicit criteria, most existing pipelines remain pointwise: they score each response independently and derive preferences by comparing aggregated scores. We show that this design is poorly matched to pairwise code preference prediction and can underperform a strong monolithic judge. We propose CriterAlign, a criterion-centric framework that adapts rubric-based judging to pairwise preference evaluation through direct criterion-level pairwise judgments, tie-driven criterion refinement, swap-consistency filtering, and final pairwise synthesis. We further introduce Human-Preference-Aligned Guidance (HPAG), synthesized offline from training examples by extracting recurring rationale gaps between human preferences and monolithic judge predictions, and injected into the criterion generator, criterion judge, and final judge. On BigCodeReward, CriterAlign improves a Qwen2.5-VL-32B monolithic judge from 60.4% to 66.3% accuracy, with ablations confirming the contributions of pairwise criterion design and HPAG.

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.

While automatic metrics drive progress in Machine Translation (MT) and Text Summarization (TS), existing metrics have been developed and validated almost exclusively for English and other high-resource languages. This narrow focus leaves Indian languages, spoken by over 1.5 billion people, largely overlooked, casting doubt on the universality of current evaluation practices. To address this gap, we introduce ITEM, a large-scale benchmark that systematically evaluates the alignment of 26 automatic metrics with human judgments across six major Indian languages, enriched with fine-grained annotations. Our extensive evaluation, covering agreement with human judgments, sensitivity to outliers, language-specific reliability, inter-metric correlations, and resilience to controlled perturbations, reveals four central findings: (1) LLM-based evaluators show the strongest alignment with human judgments at both segment and system levels; (2) outliers exert a significant impact on metric-human agreement; (3) in TS, metrics are more effective at capturing content fidelity, whereas in MT, they better reflect fluency; and (4) metrics differ in their robustness and sensitivity when subjected to diverse perturbations. Collectively, these findings offer critical guidance for advancing metric design and evaluation in Indian languages.

Sycophantic agreement refers to a behavior in which language models excessively affirm the user, often at the cost of factual accuracy. Although sycophantic agreement is a well-known failure of model alignment, there is limited understanding of how it emerges from model training. In this work, we demonstrate that sycophantic agreement can emerge as an unintended consequence of widely used contrastive preference optimization objectives. Using the OLMo 3 post-training pipeline, we show that, for various pairs of teacher models across three families, there is a strong correlation between the log-ratio of the teacher model sycophantic agreement rates and the resulting student model sycophantic agreement rate. We further demonstrate that this unintended transfer is not limited to DPO but also occurs across 6 other preference optimization objectives. To understand whether this effect can be attributed to particular training examples, we analyze the preference data and find that the sycophancy signal is diffused across the entire dataset rather than concentrated in a sparse set of examples: each example appears neutral, i.e., there are no explicit instances of sycophantic agreement, and filtering based on probe-based data attribution or logit-linear selection fails to mitigate sycophancy without removing a large portion of the dataset. Overall, our findings suggest that the teacher models used to generate preference data can interact with alignment training objectives in unexpected ways, generalizing to undesirable and potentially harmful behaviors like sycophantic agreement.

Large language models (LLMs) are entering legal workflows, yet we lack a jurisdiction-specific framework to assess their baseline competence therein. We use India's public legal examinations as a transparent proxy. Our multi-year benchmark assembles objective screens from top national and state exams and evaluates open and frontier LLMs under real-world exam conditions. To probe beyond multiple-choice questions, we also include a lawyer-graded, paired-blinded study of long-form answers from the Supreme Court's Advocate-on-Record exam. This is, to our knowledge, the first exam-grounded, India-specific yardstick for LLM court-readiness released with datasets and protocols. Our work shows that while frontier systems consistently clear historical cutoffs and often match or exceed recent top-scorer bands on objective exams, none surpasses the human topper on long-form reasoning. Grader notes converge on three reliability failure modes: procedural or format compliance, authority or citation discipline, and forum-appropriate voice and structure. These findings delineate where LLMs can assist (checks, cross-statute consistency, statute and precedent lookups) and where human leadership remains essential: forum-specific drafting and filing, procedural and relief strategy, reconciling authorities and exceptions, and ethical, accountable judgment.

Recent advancements in Language Models (LMs) have catalyzed the creation of multiple benchmarks, designed to assess these models' general capabilities. A crucial task, however, is assessing the validity of the benchmarks themselves. This is most commonly done via Benchmark Agreement Testing (BAT), where new benchmarks are validated against established ones using some agreement metric (e.g., rank correlation). Despite the crucial role of BAT for benchmark builders and consumers, there are no standardized procedures for such agreement testing. This deficiency can lead to invalid conclusions, fostering mistrust in benchmarks and upending the ability to properly choose the appropriate benchmark to use. By analyzing over 40 prominent benchmarks, we demonstrate how some overlooked methodological choices can significantly influence BAT results, potentially undermining the validity of conclusions. To address these inconsistencies, we propose a set of best practices for BAT and demonstrate how utilizing these methodologies greatly improves BAT robustness and validity. To foster adoption and facilitate future research,, we introduce BenchBench, a python package for BAT, and release the BenchBench-leaderboard, a meta-benchmark designed to evaluate benchmarks using their peers. Our findings underscore the necessity for standardized BAT, ensuring the robustness and validity of benchmark evaluations in the evolving landscape of language model research. BenchBench Package: https://github.com/IBM/BenchBench Leaderboard: https://huggingface.co/spaces/per/BenchBench

As agentic systems tackle increasingly complex multi-step tasks, evaluating their trajectories presents a major bottleneck - human annotation of a single trajectory on popular agentic benchmarks can take hours, making it difficult to scale evaluations for measuring performance or curating training data. This has driven widespread reliance on automated approaches such as LLM-as-a-judge (LLMJ) to critique agents at the process and outcome-levels at scale, however, the soundness of LLMJ critiques often goes unmeasured. Here, we introduce Counsel, the first public dataset of meta-evaluations for agentic tasks. Counsel consists of process-level critiques from open-weight LLMJs on two agent benchmarks: tau-bench (customer support agents) and DA-Code (coding agents), and human meta-evaluations of these critiques. Human annotators label critiques on each flagged error as "spot on", "correct location but poor reasoning", or "should not have flagged", achieving reliable inter-annotator agreement (Krippendorff's alpha of 0.78). The resulting dataset stratifies LLMJ critiques by human alignment across both error location within a trajectory and reasoning quality, serving as valuable data to calibrate, improve, or train LLMJs for agents. Comparing open-weight judges, we find that more capable judge models and more reasoning effort both enabled improved human agreement, with the strongest judge reaching ~88% agreement on location and ~65% on reasoning. Counsel is generated using open-weight models and is permissively licensed for broad community use, which we hope will enable rigorous study and improved alignment of LLM-based evaluators for agentic systems.

Agentic coding assistants are increasingly given extra capabilities, such as browser based testing tools and design oriented system prompts, on the assumption that more capability yields better software. This study tested that assumption directly. Ninety independent agent runs built the same application, a real time retrospective board, from one detailed specification, each scored on a fixed 14 criterion functional rubric (42 point maximum) and a visual quality review. The runs spanned several model generations, two agent harnesses, two reasoning effort levels, a testing tool, and two design oriented prompts. Capability tier dominated: frontier models clustered near the ceiling while a low cost local model fell to 24 to 37 points. A criterion level analysis revealed what run totals conceal. Container deployment was the dominant defect, failing first try in 44 percent of runs, with its failure rate shifting sharply across model generations while mean totals moved less than a point. The testing tool raised cost by 42 to 68 percent without improving functional score or reliability, even on interface visible criteria. Raising reasoning effort from High to xHigh lifted first try perfect runs from 28 percent to 89 percent and cut corrective prompts about five fold, for 9 to 29 percent more cost. A design oriented prompt raised visual quality, 4.5 versus 3.0 on a 5 point scale, without lifting function, and a one paragraph paraphrase of its directive reproduced the entire lift. The practical lesson is to match the fix to the failure: most first run failures came from weak reasoning, which a stronger model or more effort prevents, not from visible flaws a checking tool would catch.

Offering a promising solution to the scalability challenges associated with human evaluation, the LLM-as-a-judge paradigm is rapidly gaining traction as an approach to evaluating large language models (LLMs). However, there are still many open questions about the strengths and weaknesses of this paradigm, and what potential biases it may hold. In this paper, we present a comprehensive study of the performance of various LLMs acting as judges. We leverage TriviaQA as a benchmark for assessing objective knowledge reasoning of LLMs and evaluate them alongside human annotations which we found to have a high inter-annotator agreement. Our study includes 9 judge models and 9 exam taker models -- both base and instruction-tuned. We assess the judge model's alignment across different model sizes, families, and judge prompts. Among other results, our research rediscovers the importance of using Cohen's kappa as a metric of alignment as opposed to simple percent agreement, showing that judges with high percent agreement can still assign vastly different scores. We find that both Llama-3 70B and GPT-4 Turbo have an excellent alignment with humans, but in terms of ranking exam taker models, they are outperformed by both JudgeLM-7B and the lexical judge Contains, which have up to 34 points lower human alignment. Through error analysis and various other studies, including the effects of instruction length and leniency bias, we hope to provide valuable lessons for using LLMs as judges in the future.

When many people highlight the same document, is the crowd a single consensus, or is it internally structured into reader sub-groups that mark different things -- and is that structure a stable property of a reader or of the document? Building on prior work showing an individual's within-document highlighting signal is a whisper while individuality lives in selection, we ask the group-level question on a co-readership platform using a margin-preserving curveball null. Experiment 1: within a document, readers form strong sub-groups -- pairs agree far beyond what shared salience, mark density, and sentence popularity predict (nearest-neighbour agreement z=+6.3, significant in 88% of documents). Under an eight-block region-preserving null, shared engagement with the same coarse regions of the document accounts for about 40% of this excess; the majority survives as finer reader-specific agreement (z=+3.6, 77% significant). So the within-document crowd is, in a descriptive sense, factional. Experiment 2: is that grouping a stable reader trait? Here we are honest about power. The cross-document split-half reproducibility of a pair's agreement is near zero pooled (+0.078 and 0.000 in two separately drawn samples), and a power calibration shows the test is informative only for pairs that co-read many documents. In the only informative high-overlap subset (k>=4), point estimates are positive but small-sample, imprecise across the separately drawn samples, never significant, and attenuate under the region-preserving null. We therefore leave cross-document stability unresolved: the data is consistent with anything from situational grouping to a weak-to-moderate stable reader trait. The crowd is factional within a document; whether its factions follow the reader across documents is, honestly, beyond our reach.

LLM-as-a-Judge has been widely applied to evaluate and compare different LLM alignmnet approaches (e.g., RLHF and DPO). However, concerns regarding its reliability have emerged, due to LLM judges' biases and inconsistent decision-making. Previous research has developed evaluation frameworks to assess reliability of LLM judges and their alignment with human preferences. However, the employed evaluation metrics often lack adequate explainability and fail to address LLM internal inconsistency. Additionally, existing studies inadequately explore the impact of various prompt templates when applying LLM-as-a-Judge methods, leading to potentially inconsistent comparisons between different alignment algorithms. In this work, we systematically evaluate LLM-as-a-Judge on alignment tasks by defining more theoretically interpretable evaluation metrics and explicitly mitigating LLM internal inconsistency from reliability metrics. We develop an open-source framework to evaluate, compare, and visualize the reliability and alignment of LLM judges, which facilitates practitioners to choose LLM judges for alignment tasks. In the experiments, we examine effects of diverse prompt templates on LLM-judge reliability and also demonstrate our developed framework by comparing various LLM judges on two common alignment datasets (i.e., TL;DR Summarization and HH-RLHF-Helpfulness). Our results indicate a significant impact of prompt templates on LLM judge performance, as well as a mediocre alignment level between the tested LLM judges and human evaluators.

Disagreement in annotation is a common phenomenon in the development of NLP datasets and serves as a valuable source of insight. While majority voting remains the dominant strategy for aggregating labels, recent work has explored modeling individual annotators to preserve their perspectives. However, modeling each annotator is resource-intensive and remains underexplored across various NLP tasks. We propose an agreement-based clustering technique to model the disagreement between the annotators. We conduct comprehensive experiments in 40 datasets in 18 typologically diverse languages, covering three subjective NLP tasks: sentiment analysis, emotion classification, and hate speech detection. We evaluate four aggregation approaches: majority vote, ensemble, multi-label, and multitask. The results demonstrate that agreement-based clustering can leverage the full spectrum of annotator perspectives and significantly enhance classification performance in subjective NLP tasks compared to majority voting and individual annotator modeling. Regarding the aggregation approach, the multi-label and multitask approaches are better for modeling clustered annotators than an ensemble and model majority vote.

This research introduces the Judge's Verdict Benchmark, a novel two-step methodology to evaluate Large Language Models (LLMs) as judges for response accuracy evaluation tasks. We assess how well 54 LLMs can replicate human judgment when scoring responses from RAG (Retrieval-Augmented Generation) or Agentic pipelines against ground truth answers. Our methodology progresses from traditional correlation analysis to comprehensive Cohen's Kappa analysis that measures actual agreement patterns. The two-step approach includes: (1) a correlation test that filters judges with strong alignment, followed by (2) a human-likeness test using z-scores to identify two distinct judgment patterns: human-like judgment (|z| < 1) that mimics natural human variation, and super-consistent judgment (z > 1) that exceeds typical human-to-human agreement levels. This methodology reveals that 27 out of 54 tested LLMs achieve Tier 1 performance: 23 models exhibit human-like patterns that preserve the nuances of human judgment, while 4 models demonstrate super-consistent behavior, a pattern that could indicate either enhanced reliability or oversimplification of complex judgments. Testing 43 open-source models (1B-405B parameters) and 11 closed models (GPT, Gemini, Claude variants), we demonstrate that judge excellence is not solely dependent on model size but on specific training strategies. Our key contributions include: (1) establishing that correlation alone is insufficient for judge evaluation, (2) introducing a "Turing Test for judges" based on agreement patterns, and (3) providing a standardized benchmark for classifying LLM judges into distinct performance tiers for different evaluation needs.

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.

Using a large language model to play the user is now standard in scalable evaluation. It has a repeatedly diagnosed failure: the simulated user is excessively cooperative, so a system under test can score by the sheer number of questions it asks rather than by making the user willing to speak. We answer with a disclosure gate conditioning information release on the companion agent's behaviour: its state is a ladder of five ordered gates, merged onto three observable depth layers. We specify, ablate, and audit it, and train a user simulator against that specification. Gating behaviour is learned from the training corpus's synthetic branch, while the real branch supplies how people speak and react; after training, the simulator need not be told at runtime which gate each item sits behind. The gate is a load-bearing component of the environment: on the English corpus of a published companion-agent benchmark (CompanionBench), once training no longer states per example which gate each item sits behind, the largest rank displacement across 12 systems under test exceeds the noise band set by re-running that environment under a new seed, while per-system scores show no detectable change. We state two acceptance criteria: a ranking must be order-preserving, and absolute scores must be scale-stable. Of the candidates we examine, only one passes both -- the simulator we release -- and its leaderboard correlates at 0.993 with the benchmark's original simulator. By contrast, prompting a frontier model as the simulator barely moves the ranking while shifting every score upward -- a shift invisible to anyone checking the ranking alone. The environment we specify is the one that benchmark already used. That publication describes the mechanism in about four hundred words, and we supply what it lacked: specification, ablations, human studies, negative controls, and downstream sensitivity analysis.

Matching submissions with suitable reviewers at scale is a growing challenge for major venues, yet existing approaches either rely on coarse proxy signals that conflate general relatedness with true suitability, or require expensive human annotations that are difficult to scale for training. We propose MERIT, a two-stage framework that bridges this gap by converting criterion-level expertise matching into scalable suitability supervision. In the first stage, we train a reviewer assessor via reinforcement learning to identify the expertise dimensions a paper requires, match them against the reviewer's prior work, and produce a suitability decision, with rewards provided by an LLM judge guided by paper-specific expertise rubrics. In the second stage, we distill the assessor's predictions into an embedding-based retriever for efficient large-scale assignment. Experiments show that our 4B reviewer assessor outperforms larger general-purpose LLMs on suitability classification, and the resulting retriever achieves state-of-the-art performance across LR-Bench and the CMU Gold dataset. Our code is available at https://github.com/Luli3220/MERIT.

Professional-document review is a constraint-checking problem in which decisions depend on relations among text, geometry, pages, and document revisions. We present an evidence-grounded pipeline that normalizes extracted facts, executes four-state rules deterministically, retains source spans, and escalates unresolved cases. We evaluate its PDF evidence allocator on 160 reference-based tasks from 29 construction projects using a repeated four-system test and a disjoint two-system breadth extension. In the repeated test, reallocating a four-image budget from retrieved page overviews to one overview and three overlapping tiles improves project-family standardized decision accuracy by 10.6 percentage points (95% project-cluster bootstrap CI: 4.3 to 18.0; exact p = 0.031). This effect does not persist in the broader block: Region-RAG changes accuracy by -4.1 points (95% CI: -10.2 to 1.9; exact p = 0.209), while an equal-image sensitivity favors page breadth. Exact finding-set recovery remains low, false passes remain common, and repeated-run agreement is poorly calibrated. The results identify a resolution-breadth trade-off rather than a universal advantage for region-focused evidence, motivating rule-aware evidence routing and expert review.

Evaluating generative video models remains an open problem. Reference-based metrics such as Structural Similarity Index Measure (SSIM) and Peak Signal to Noise Ratio (PSNR) reward pixel fidelity over semantic correctness, while Frechet Video Distance (FVD) favors distributional textures over physical plausibility. Binary Visual Question Answering (VQA) based benchmarks like VBench~2.0 are prone to yes-bias and rely on low-resolution auditors that miss temporal failures. Moreover, their prompts target a single dimension at a time, multiplying the number of videos required while still not guaranteeing reliable results. WorldJen addresses these limitations directly. Binary VQA is replaced with Likert-scale questionnaires graded by a VLM that receives frames at native video resolution. Video generation costs are addressed by using adversarially curated prompts that are designed to exercise up to 16 quality dimensions simultaneously. The framework is built around two interlocking contributions. First, A blind human preference study is conducted, accumulating (2,696 pairwise annotations from 7 annotators with 100% pair coverage over 50 of the curated prompts times 6 state-of-the-art video models. A mean inter-annotator agreement of 66.9% is achieved and the study establishes a human ground-truth Bradley-Terry (BT) rating with a three-tier structure. Second, A VLM-as-a-judge evaluation engine using prompt-specific, dimension-specific Likert questionnaires (10 questions per dimension, 47,160 scored responses) judges the videos and reproduces the human-established three-tier BT rating structure independently. The VLM achieves a Spearman hatρ=1.000,~p=0.0014 that is interpreted as tier agreement with the human results. Six focused ablation studies validate the robustness of the VLM evaluation framework.

Current approaches to AI compliance treat conformity as a binary, audit-time verdict rather than a continuous, measurable property of production systems. We argue that this compliance fiction is structurally ill-suited to the requirements of the EU AI Act, which demands ongoing human oversight and the detection of emergent behavioural drift in deployed systems. We introduce governance from metrics, a principle whereby regulatory compliance is derived as a continuous signal from runtime observability rather than from static assessments. Building on this principle, we present govllm, an open-source framework implementing a governance-driven routing architecture in which model selection is determined by accumulated compliance scores rather than by latency or cost alone. Central to our approach is a panel of regulatory judges - LLM evaluators specialised per criterion (EU AI Act, GDPR, ANSSI, accessibility) - whose inter-judge disagreement we reframe not as noise but as a regulatory uncertainty signal warranting human arbitration. We validate this approach through a ground truth corpus of 49 annotated prompt/response pairs across five regulatory criteria, evaluated by four small language models (SLMs, 1.7B-7B parameters) running fully on-premise. Agreement rates range from 51.5% (mistral:7b) to 69.1% (phi4-mini), with no single model dominating across all criteria - empirically motivating the Profile-as-jury design. We further document three structural failure modes in small regulatory judges and a judge-specific position bias that degrades agreement by up to 25 percentage points across three question-order conditions (original, reversed, permuted). govllm is released as open-source software to support reproducible AI governance research.

Disagreements are common in online discussions. Disagreement may foster collaboration and improve the quality of a discussion under some conditions. Although there exist methods for recognizing disagreement, a deeper understanding of factors that influence disagreement is lacking in the literature. We investigate a hypothesis that differences in personal values are indicative of disagreement in online discussions. We show how state-of-the-art models can be used for estimating values in online discussions and how the estimated values can be aggregated into value profiles. We evaluate the estimated value profiles based on human-annotated agreement labels. We find that the dissimilarity of value profiles correlates with disagreement in specific cases. We also find that including value information in agreement prediction improves performance.

Agentic retrieval-augmented reasoning pipelines are increasingly used to structure how large language models (LLMs) incorporate external evidence in clinical decision support. These systems iteratively retrieve curated domain knowledge and synthesize it into structured reports before answer selection. Although such pipelines can improve performance, their impact on reliability under model variability remains unclear. In real-world deployment, heterogeneous models may align, diverge, or synchronize errors in ways not captured by accuracy. We evaluated 34 LLMs on 169 expert-curated publicly available radiology questions, comparing zero-shot inference with a radiology-specific multi-step agentic retrieval condition in which all models received identical structured evidence reports derived from curated radiology knowledge. Agentic inference reduced inter-model decision dispersion (median entropy 0.48 vs. 0.13) and increased robustness of correctness across models (mean 0.74 vs. 0.81). Majority consensus also increased overall (P<0.001). Consensus strength and robust correctness remained correlated under both strategies (ho=0.88 for zero-shot; ho=0.87 for agentic), although high agreement did not guarantee correctness. Response verbosity showed no meaningful association with correctness. Among 572 incorrect outputs, 72% were associated with moderate or high clinically assessed severity, although inter-rater agreement was low (appa=0.02). Agentic retrieval therefore was associated with more concentrated decision distributions, stronger consensus, and higher cross-model robustness of correctness. These findings suggest that evaluating agentic systems through accuracy or agreement alone may not always be sufficient, and that complementary analyses of stability, cross-model robustness, and potential clinical impact are needed to characterize reliability under model variability.

Automated L2 speech assessment can assign proficiency labels, but often lacks interpretability. We propose a rubric-guided SpeechLLM for multi-aspect, multi-granular assessment, trained with a hybrid objective combining supervised fine-tuning and Bounded Direct Preference Optimization. The model jointly predicts ordinal labels at the sentence-level (accuracy, fluency, prosody), word/phoneme-level accuracy, and generates a natural-language rationale in the same response. On SpeechOcean762, our approach matches or outperforms single-granularity models while remaining competitive with prior approaches. We analyze rationale reliability along two axes: self-consistency with model predictions and alignment with ground-truth labels, using sentiment consistency (plausibility) and mention-based agreement (faithfulness). Rationales are plausible at the sentence level, but faithfulness degrades at the word/phoneme level: references are sparse and weakly aligned with token-level labels.

Large language models are increasingly used for social decision-making situations that require balancing cultural norms with personal preferences. For example, a user preferring honesty might ask whether to correct a coworker publicly when local norms favor indirect feedback. Yet existing research studies cultural alignment and personalization largely separately. We introduce PACT, the Personal-Preference and Cultural-Norm Trade-off framework, which evaluates whether models choose to follow a cultural norm or allow personal preferences. We find that LLMs vary in how rigidly they enforce cultural norms, with behavior shifted more by country context (7.8%) than age (1%) and gender (0.7%) and shifting non-uniformly after instruction tuning. Furthermore, our five-country human study on PACT shows that culture-following in humans is mainly driven by scenario country, with the lowest agreement when participants judge their own cultural contexts, showing within-culture pluralism. Finally, human-LLM alignment experiments show that models can match majority choices, but fail to capture response distributions and uncertainty (with best correlations reaching only 0.24). Together, these findings motivate alignment evaluations that go beyond majority to capture cultural pluralism and disagreement in social judgment.

Per-field accept/review with selective risk at most alpha -- accept a field only if the error rate among accepted fields is controlled -- is the trust contract document-extraction systems need, and the natural procedure silently violates it on real documents. On 13,859 genuine claude-sonnet-5 fields from 800 CORD receipts (49.0% correct) we diagnose three failure modes: document clustering (design effect 1.84-2.45), score-refit leakage (coverage 0.416 at risk 0.127, violating alpha=0.10 in 95% of splits), and a tie-mass pathology (a degenerate score collapses the threshold grid, 0.030 to 0.001). We organize the fixes as a validity ladder, guarantee form stated per tier. A fit/val split protocol restores expected-selective-risk control for a learned fusion: coverage 0.318 at risk 0.096 at nominal alpha=0.10, no tolerance band (production variant 0.326) -- an on-average point whose realized risk exceeds alpha in 47.5% of resplits, not a certificate. Mondrian Learn-then-Test with exact binomial tails yields per-group PAC certificates: field-iid 0.171 at risk 0.068, cluster-corrected 0.140, doc-iid 0.060 -- the only tier matching documents, honestly near-vacuous today. Support-bin, the pre-specified provenance taxonomy, wins every rigor tier on the sonnet CORD capture (p<1e-4, Bonferroni-corrected) -- a win that does not replicate on the same documents under haiku or qwen -- while on higher-accuracy corpora pooled thresholds win: conditioning helps exactly where pooled cannot certify, subsumed by a learned score elsewhere. A frozen-configuration confirmation on selection-untouched claude-haiku-4-5 held at both risk levels, and a blind three-annotator human-gold audit verifies the practical tier's accepted-set risk at 1.3% against its 10% budget (Fleiss' kappa=0.83; labels err one-sidedly pessimistic). Released Apache-2.0 with seed-pinned, regression-gated procedures.

Text-to-image models now generate graphic design at production scale, yet their supervision still comes primarily from photo-style preference datasets with a single overall verdict per comparison. Designers evaluate designs along several distinct axes (e.g., typography, layout, color harmony) that a single preference label collapses. We release TASTE (Typography, Aesthetics, Spatial, Tone, Etc.), a multi-dimensional preference dataset in which two disjoint cohorts of five professional designers each ranked outputs from four current text-to-image models across nine criteria along with per-image hallucination flags. We pair the dataset with two contributions. First, a criterion-agnostic signal-validation framework based on Kendall's τ, majority-vote probability, and Condorcet cycles against exact iid-uniform nulls; the analysis reveals significant but moderate designer agreement, with every TASTE criterion rejecting the random-rater null. Second, we benchmark preference models on TASTE and find that off-the-shelf VLM judges and dedicated T2I scorers fail to reach majority agreement with the designer panel, while a small MLP head trained directly on TASTE substantially narrows the gap to the single-rater ceiling, setting a baseline for future TASTE-trained preference models.

Radiological imaging is central to diagnosis, treatment planning, and clinical decision-making. Vision-language foundation models have spurred interest in automated radiology report generation (RRG), but safe deployment requires reliable clinical evaluation of generated reports. Existing metrics often rely on surface-level similarity or behave as black boxes, lacking interpretability. We introduce ICARE (Interpretable and Clinically-grounded Agent-based Report Evaluation), an interpretable evaluation framework leveraging large language model agents and dynamic multiple-choice question answering (MCQA). Two agents, each with either the ground-truth or generated report, generate clinically meaningful questions and quiz each other. Agreement on answers captures preservation and consistency of findings, serving as interpretable proxies for clinical precision and recall. By linking scores to question-answer pairs, ICARE enables transparent, and interpretable assessment. Clinician studies show ICARE aligns significantly more with expert judgment than prior metrics. Perturbation analyses confirm sensitivity to clinical content and reproducibility, while model comparisons reveal interpretable error patterns.

Despite the subjective nature of many NLP tasks, most NLU evaluations have focused on using the majority label with presumably high agreement as the ground truth. Less attention has been paid to the distribution of human opinions. We collect ChaosNLI, a dataset with a total of 464,500 annotations to study Collective HumAn OpinionS in oft-used NLI evaluation sets. This dataset is created by collecting 100 annotations per example for 3,113 examples in SNLI and MNLI and 1,532 examples in Abductive-NLI. Analysis reveals that: (1) high human disagreement exists in a noticeable amount of examples in these datasets; (2) the state-of-the-art models lack the ability to recover the distribution over human labels; (3) models achieve near-perfect accuracy on the subset of data with a high level of human agreement, whereas they can barely beat a random guess on the data with low levels of human agreement, which compose most of the common errors made by state-of-the-art models on the evaluation sets. This questions the validity of improving model performance on old metrics for the low-agreement part of evaluation datasets. Hence, we argue for a detailed examination of human agreement in future data collection efforts, and evaluating model outputs against the distribution over collective human opinions. The ChaosNLI dataset and experimental scripts are available at https://github.com/easonnie/ChaosNLI

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.

Evaluating Large Language Models (LLMs) for mental health support is challenging due to the emotionally and cognitively complex nature of therapeutic dialogue. Existing benchmarks are limited in scale, reliability, often relying on synthetic or social media data, and lack frameworks to assess when automated judges can be trusted. To address the need for large-scale dialogue datasets and judge reliability assessment, we introduce two benchmarks that provide a framework for generation and evaluation. MentalBench-100k consolidates 10,000 one-turn conversations from three real scenarios datasets, each paired with nine LLM-generated responses, yielding 100,000 response pairs. MentalAlign-70k}reframes evaluation by comparing four high-performing LLM judges with human experts across 70,000 ratings on seven attributes, grouped into Cognitive Support Score (CSS) and Affective Resonance Score (ARS). We then employ the Affective Cognitive Agreement Framework, a statistical methodology using intraclass correlation coefficients (ICC) with confidence intervals to quantify agreement, consistency, and bias between LLM judges and human experts. Our analysis reveals systematic inflation by LLM judges, strong reliability for cognitive attributes such as guidance and informativeness, reduced precision for empathy, and some unreliability in safety and relevance. Our contributions establish new methodological and empirical foundations for reliable, large-scale evaluation of LLMs in mental health. We release the benchmarks and codes at: https://github.com/abeerbadawi/MentalBench/

Automatic dialogue coherence evaluation has attracted increasing attention and is crucial for developing promising dialogue systems. However, existing metrics have two major limitations: (a) they are mostly trained in a simplified two-level setting (coherent vs. incoherent), while humans give Likert-type multi-level coherence scores, dubbed as "quantifiable"; (b) their predicted coherence scores cannot align with the actual human rating standards due to the absence of human guidance during training. To address these limitations, we propose Quantifiable Dialogue Coherence Evaluation (QuantiDCE), a novel framework aiming to train a quantifiable dialogue coherence metric that can reflect the actual human rating standards. Specifically, QuantiDCE includes two training stages, Multi-Level Ranking (MLR) pre-training and Knowledge Distillation (KD) fine-tuning. During MLR pre-training, a new MLR loss is proposed for enabling the model to learn the coarse judgement of coherence degrees. Then, during KD fine-tuning, the pretrained model is further finetuned to learn the actual human rating standards with only very few human-annotated data. To advocate the generalizability even with limited fine-tuning data, a novel KD regularization is introduced to retain the knowledge learned at the pre-training stage. Experimental results show that the model trained by QuantiDCE presents stronger correlations with human judgements than the other state-of-the-art metrics.

Many math- and science-oriented agent systems use hierarchical designs with specialized reviewer roles, assuming that a dedicated review stage should help turn wrong candidates into correct ones. We test this assumption on 4,181 verifier-grounded Omni-MATH problems using matched gpt-oss-120b actors. Collaboration adds little on the easiest tiers, but from tier 4 onward the gains open sharply; in this harder regime, broadcast-style peer discussion reaches higher final accuracy than a planner-executor-reviewer pipeline (PER). We ask whether this gap is explained by reviewer quality or by whether critique changes the next answer the protocol carries forward. It is not explained by reviewer precision alone: PER's reviewer is more precise than broadcast's (0.861 vs. 0.644), yet evaluator-verified useful critique is much less likely to change the next candidate and produces lower reviewer-guided repair. These results show that reviewer detection quality and critique uptake are empirically separable. Within matched PER interventions, forcing explicit acknowledgment lowers final accuracy, while embedding reviewer guidance directly in the solver's working context partially improves follow-through without closing the gap. Overall, reviewer-centric evaluation can overstate system quality: a protocol may spot errors well yet still fail to solve more problems if it does not act on those critiques.

As large language models increasingly participate in scientific evaluation, we investigate a potential form of reward hacking: how rhetorical choices shape AI-review judgments when reported scientific content is preserved and how these effects vary across evaluation conditions. We construct a controlled corpus of 4,200 full-paper manuscripts derived from 120 anonymized ICLR 2026 submissions. Two LLM rewriters transform six rhetorical dimensions in opposing directions, and five LLM reviewers evaluate the resulting manuscripts under standard and strict protocols. We also test joint, recursive, and reviewer-guided rewriting. Our results show that rhetorical sensitivity is structured rather than uniform. Evidence framing and novelty stance produce the largest positive-negative contrasts in overall assessment, with scope framing forming a weaker second tier; the remaining dimensions have smaller or less stable effects. This hierarchy persists across human-assessed quality levels, but score movement depends strongly on the AI reviewer's original score: lower scores tend to rise, higher scores tend to fall, and directional contrasts are clearest in the middle ranges. More elaborate workflows do not reliably yield larger gains. Joint rewriting is strongly rewriter-dependent, reviewer guidance does not consistently outperform an unguided second pass, and repeated rewriting yields diminishing, configuration-dependent returns. Across conditions, the rewriter primarily determines the separation between opposing variants, whereas the reviewer determines the magnitude and sign of their score effects. Strict review lowers mean OA by 1.36 points without consistently changing rhetorical sensitivity. These findings identify when rhetorical presentation influences AI scientific review and motivate evaluation systems robust to content-preserving variation in scientific writing.

In Natural Language Processing (NLP), the Elo rating system, originally designed for ranking players in dynamic games such as chess, is increasingly being used to evaluate Large Language Models (LLMs) through "A vs B" paired comparisons. However, while popular, the system's suitability for assessing entities with constant skill levels, such as LLMs, remains relatively unexplored. We study two fundamental axioms that evaluation methods should adhere to: reliability and transitivity. We conduct extensive evaluation of Elo behaviour, illustrating that individual Elo computations exhibit volatility and delving into the impact of varying the Elo rating system's hyperparameters. We show that these axioms are not always satisfied raising questions about the reliability of current comparative evaluations of LLMs. If the current use of Elo scores is intended to substitute the costly head-to-head comparison of LLMs, it is crucial to ensure the ranking is as robust as possible. Guided by the axioms, our findings offer concrete guidelines for enhancing the reliability of LLM evaluation methods, suggesting a need for reassessment of existing comparative approaches.

Large language models excel on objectively verifiable tasks such as math and programming, where evaluation reduces to unit tests or a single correct answer. In contrast, real-world enterprise work is often subjective and context-dependent: success hinges on organizational goals, user intent, and the quality of intermediate artifacts produced across long, multi-tool workflows. We introduce LH-Bench, a three-pillar evaluation design that moves beyond binary correctness to score autonomous, long-horizon execution on subjective enterprise tasks. The pillars are: (i) expert-grounded rubrics that give LLM judges the domain context needed to score subjective work, (ii) curated ground-truth artifacts that enable stepwise reward signals (e.g., chapter-level annotation for content tasks), and (iii) pairwise human preference evaluation for convergent validation. We show that domain-authored rubrics provide substantially more reliable evaluation signals than LLM-authored rubrics (kappa = 0.60 vs. 0.46), and that human preference judgments confirm the same top-tier separation (p < 0.05), evidence that expert-grounded evaluation can scale without sacrificing reliability. We release public datasets and report results on two environments: Figma-to-code (33 real .fig tasks against the Figma API via MCP) and Programmatic content (41 courses comprising 183 individually-evaluated chapters on a course platform serving 30+ daily users).

Reinforcement learning with verifiable rewards has made post-training highly effective when correctness can be checked automatically. However, many important model behaviors require satisfying several qualitative criteria at once. Rubric-based rewards address this setting by grading prompt-specific criteria and aggregating them into a scalar reward. Yet standard static aggregations conflate a criterion's human-assigned importance with its current usefulness as an optimization signal. We show that this assumption breaks down in rubric RL: many important criteria are already saturated or currently unreachable, while criteria that distinguish rollouts are not necessarily those with the largest human weights. We introduce POW3R, a policy-aware rubric reward framework that preserves human weights and category balance as the rubric objective while adapting criterion-level reward weights during training. POW3R uses rollout-level contrast to emphasize criteria that currently separate the policy's outputs, making the GRPO reward more informative without changing the underlying evaluation target. Across three base policies on two datasets spanning multimodal and text-only settings, POW3R wins 24 of 30 base-policy/metric comparisons, improving both mean rubric reward and strict completion (the fraction of prompts whose response satisfies every required rubric criterion) over vanilla GRPO with rubric rewards, and reaches the same plateau in 2.5--4times fewer training steps. Rubric rewards should therefore distinguish what should matter in the final answer from what can teach the current policy.

Language model heavily depends on high-quality data for optimal performance. Existing approaches rely on manually designed heuristics, the perplexity of existing models, training classifiers, or careful prompt engineering, which require significant expert experience and human annotation effort while introduce biases. We introduce CritiQ, a novel data selection method that automatically mines criteria from human preferences for data quality with only sim30 human-annotated pairs and performs efficient data selection. The main component, CritiQ Flow, employs a manager agent to evolve quality criteria and worker agents to make pairwise judgments. We build a knowledge base that extracts quality criteria from previous work to boost CritiQ Flow. Compared to perplexity- and classifier- based methods, verbal criteria are more interpretable and possess reusable value. After deriving the criteria, we train the CritiQ Scorer to give quality scores and perform efficient data selection. We demonstrate the effectiveness of our method in the code, math, and logic domains, achieving high accuracy on human-annotated test sets. To validate the quality of the selected data, we continually train Llama 3.1 models and observe improved performance on downstream tasks compared to uniform sampling. Ablation studies validate the benefits of the knowledge base and the reflection process. We analyze how criteria evolve and the effectiveness of majority voting.

Evaluating free-text explanations is a multifaceted, subjective, and labor-intensive task. Large language models (LLMs) present an appealing alternative due to their potential for consistency, scalability, and cost-efficiency. In this work, we present ACORN, a new dataset of 3,500 free-text explanations and aspect-wise quality ratings, and use it to gain insights into how LLMs evaluate explanations. We observed that replacing one of the human ratings sometimes maintained, but more often lowered the inter-annotator agreement across different settings and quality aspects, suggesting that their judgments are not always consistent with human raters. We further quantified this difference by comparing the correlation between LLM-generated ratings with majority-voted human ratings across different quality aspects. With the best system, Spearman's rank correlation ranged between 0.53 to 0.95, averaging 0.72 across aspects, indicating moderately high but imperfect alignment. Finally, we considered the alternative of using an LLM as an additional rater when human raters are scarce, and measured the correlation between majority-voted labels with a limited human pool and LLMs as an additional rater, compared to the original gold labels. While GPT-4 improved the outcome when there were only two human raters, in all other observed cases, LLMs were neutral to detrimental when there were three or more human raters. We publicly release the dataset to support future improvements in LLM-in-the-loop evaluation here: https://github.com/a-brassard/ACORN.

Large language models (LLMs) have witnessed rapid advancements, demonstrating remarkable capabilities. However, a notable vulnerability persists: LLMs often uncritically accept flawed or contradictory premises, leading to inefficient reasoning and unreliable outputs. This emphasizes the significance of possessing the Premise Critique Ability for LLMs, defined as the capacity to proactively identify and articulate errors in input premises. Most existing studies assess LLMs' reasoning ability in ideal settings, largely ignoring their vulnerabilities when faced with flawed premises. Thus, we introduce the Premise Critique Bench (PCBench), designed by incorporating four error types across three difficulty levels, paired with multi-faceted evaluation metrics. We conducted systematic evaluations of 15 representative LLMs. Our findings reveal: (1) Most models rely heavily on explicit prompts to detect errors, with limited autonomous critique; (2) Premise critique ability depends on question difficulty and error type, with direct contradictions being easier to detect than complex or procedural errors; (3) Reasoning ability does not consistently correlate with the premise critique ability; (4) Flawed premises trigger overthinking in reasoning models, markedly lengthening responses due to repeated attempts at resolving conflicts. These insights underscore the urgent need to enhance LLMs' proactive evaluation of input validity, positioning premise critique as a foundational capability for developing reliable, human-centric systems. The code is available at https://github.com/MLGroupJLU/Premise_Critique.

Recursive self-improvement is moving from theory to practice: modern systems can critique, revise, and evaluate their own outputs, yet iterative self-modification risks subtle alignment drift. We introduce SAHOO, a practical framework to monitor and control drift through three safeguards: (i) the Goal Drift Index (GDI), a learned multi-signal detector combining semantic, lexical, structural, and distributional measures; (ii) constraint preservation checks that enforce safety-critical invariants such as syntactic correctness and non-hallucination; and (iii) regression-risk quantification to flag improvement cycles that undo prior gains. Across 189 tasks in code generation, mathematical reasoning, and truthfulness, SAHOO produces substantial quality gains, including 18.3 percent improvement in code tasks and 16.8 percent in reasoning, while preserving constraints in two domains and maintaining low violations in truthfulness. Thresholds are calibrated on a small validation set of 18 tasks across three cycles. We further map the capability-alignment frontier, showing efficient early improvement cycles but rising alignment costs later and exposing domain-specific tensions such as fluency versus factuality. SAHOO therefore makes alignment preservation during recursive self-improvement measurable, deployable, and systematically validated at scale.

While Reinforcement Learning with Verifiable Rewards (RLVR) is effective for deterministically checkable tasks, many vision-language tasks are partially verifiable, demanding multi-criteria supervision (e.g., perceptual details, reasoning steps, and constraints). Rubrics provide a natural interface for this fine-grained supervision, but their effectiveness depends on the execution accuracy during online RL. We propose Reinforcement Learning with Robust Rubric Rewards (RLR^3), extending RLVR from task-level verification to criterion-level verification. RLR^3 routes instance-specific rubrics through two execution paths: an LLM-as-an-extractor paired with a deterministic verifier, or an LLM-as-a-Judge for non-verifiable criteria. To ensure faithful scoring, RLR^3 introduce a minimal exposure strategy that masks ground truths from extractors and images from judges. Furthermore, RLR^3 employs hierarchical aggregation to prioritize essential criteria over additional criteria, and mitigates score saturation within rollout groups. Evaluated on Qwen3-VL-30B-A3B across 15 benchmarks, RLR^3 consistently outperforms RLVR, yielding a 4.7-point improvement over the base model and exceeding the official instruct-to-thinking model gap. Controlled audits confirm our deterministic verification and minimal exposure significantly reduce exploitable false positives.

The integration of Large Language Models (LLMs) into various global cultures fundamentally presents a cultural challenge: LLMs must navigate interactions, respect social norms, and avoid transgressing cultural boundaries. However, it is still unclear if LLMs can adapt their outputs to diverse cultural norms. Our study focuses on this aspect. We introduce NormAd, a novel dataset, which includes 2.6k stories that represent social and cultural norms from 75 countries, to assess the ability of LLMs to adapt to different granular levels of socio-cultural contexts such as the country of origin, its associated cultural values, and prevalent social norms. Our study reveals that LLMs struggle with cultural reasoning across all contextual granularities, showing stronger adaptability to English-centric cultures over those from the Global South. Even with explicit social norms, the top-performing model, Mistral-7b-Instruct, achieves only 81.8\% accuracy, lagging behind the 95.6\% achieved by humans. Evaluation on NormAd further reveals that LLMs struggle to adapt to stories involving gift-giving across cultures. Due to inherent agreement or sycophancy biases, LLMs find it considerably easier to assess the social acceptability of stories that adhere to cultural norms than those that deviate from them. Our benchmark measures the cultural adaptability (or lack thereof) of LLMs, emphasizing the potential to make these technologies more equitable and useful for global audiences. We release the NormAd dataset and its associated code on GitHub.

Peer review serves as a backbone of academic research, but in most AI conferences, the review quality is degrading as the number of submissions explodes. To reliably detect low-quality reviews, we define misinformed review points as either "weaknesses" in a review that contain incorrect premises, or "questions" in a review that can be already answered by the paper. We verify that 15.2% of weaknesses and 26.4% of questions are misinformed and introduce ReviewScore indicating if a review point is misinformed. To evaluate the factuality of each premise of weaknesses, we propose an automated engine that reconstructs every explicit and implicit premise from a weakness. We build a human expert-annotated ReviewScore dataset to check the ability of LLMs to automate ReviewScore evaluation. Then, we measure human-model agreements on ReviewScore using eight current state-of-the-art LLMs and verify moderate agreements. We also prove that evaluating premise-level factuality shows significantly higher agreements than evaluating weakness-level factuality. A thorough disagreement analysis further supports a potential of fully automated ReviewScore evaluation.

We present the development of a Named Entity Recognition (NER) dataset for Tagalog. This corpus helps fill the resource gap present in Philippine languages today, where NER resources are scarce. The texts were obtained from a pretraining corpora containing news reports, and were labeled by native speakers in an iterative fashion. The resulting dataset contains ~7.8k documents across three entity types: Person, Organization, and Location. The inter-annotator agreement, as measured by Cohen's kappa, is 0.81. We also conducted extensive empirical evaluation of state-of-the-art methods across supervised and transfer learning settings. Finally, we released the data and processing code publicly to inspire future work on Tagalog NLP.

We introduce PerceptionRubrics, a rubric-based evaluation framework that addresses the gap between saturated benchmark scores and real-world brittleness. Shifting evaluation from holistic semantic matching to rigorous atomic auditing, PerceptionRubrics pairs 1,038 information-dense images with over 12,000 instance-specific rubrics. These criteria are derived from golden captions constructed via a novel Circular Peer-Review consensus pipeline and then distilled into a dual-stream system of Must-Right (essential facts) and Easy-Wrong (fine-grained details) rubrics. Crucially, PerceptionRubrics implements a Gated Scoring mechanism: unlike linear averages, failure on mandatory visual facts triggers sharp binary penalties. Extensive evaluation yields critical insights: (1) The Reliability Gap: models often verify fragmented elements correctly yet fail strict conjunctive constraints, exposing brittleness in dense domains; (2) Open-Closed Stratification: contrary to reasoning trends, we reveal a persistent 8% perception deficit between open-source and proprietary frontiers; and (3) Human-Aligned Rigor: our gated metrics substantially out-align conventional benchmarks, validating that strict perceptual fidelity is the prerequisite for reliable generation.

Large language models (LLMs) and small language models (SLMs) are being adopted at remarkable speed, although their safety still remains a serious concern. With the advent of multilingual S/LLMs, the question now becomes a matter of scale: can we expand multilingual safety evaluations of these models with the same velocity at which they are deployed? To this end we introduce RTP-LX, a human-transcreated and human-annotated corpus of toxic prompts and outputs in 28 languages. RTP-LX follows participatory design practices, and a portion of the corpus is especially designed to detect culturally-specific toxic language. We evaluate seven S/LLMs on their ability to detect toxic content in a culturally-sensitive, multilingual scenario. We find that, although they typically score acceptably in terms of accuracy, they have low agreement with human judges when judging holistically the toxicity of a prompt, and have difficulty discerning harm in context-dependent scenarios, particularly with subtle-yet-harmful content (e.g. microagressions, bias). We release of this dataset to contribute to further reduce harmful uses of these models and improve their safe deployment.

Agreement Technologies refer to open computer systems in which autonomous software agents interact with one another, typically on behalf of humans, in order to come to mutually acceptable agreements. With the advance of AI systems in recent years, it has become apparent that such agreements, in order to be acceptable to the involved parties, must remain aligned with ethical principles and moral values. However, this is notoriously difficult to ensure, especially as different human users (and their software agents) may hold different value systems, i.e. they may differently weigh the importance of individual moral values. Furthermore, it is often hard to specify the precise meaning of a value in a particular context in a computational manner. Methods to estimate value systems based on human-engineered specifications, e.g. based on value surveys, are limited in scale due to the need for intense human moderation. In this article, we propose a novel method to automatically learn value systems from observations and human demonstrations. In particular, we propose a formal model of the value system learning problem, its instantiation to sequential decision-making domains based on multi-objective Markov decision processes, as well as tailored preference-based and inverse reinforcement learning algorithms to infer value grounding functions and value systems. The approach is illustrated and evaluated by two simulated use cases.

Large Language Models (LLMs) have become indispensable for evaluating writing. However, text feedback they provide is often unintelligible, generic, and not specific to user criteria. Inspired by structured rubrics in education and intelligible AI explanations, we propose iRULER following identified design guidelines to scaffold the review process by specific criteria, providing justification for score selection, and offering actionable revisions to target different quality levels. To qualify user-defined criteria, we recursively used iRULER with a rubric-of-rubrics to iteratively refine rubrics. In controlled experiments on writing revision and rubric creation, iRULER most improved validated LLM-judged review scores and was perceived as most helpful and aligned compared to read-only rubric and text-based LLM feedback. Qualitative findings further support how iRULER satisfies the design guidelines for user-defined feedback. This work contributes interactive rubric tools for intelligible LLM-based review and revision of writing, and user-defined rubric creation.

The reliance on translated or adapted datasets from English or multilingual resources introduces challenges regarding linguistic and cultural suitability. This study addresses the need for robust and culturally appropriate benchmarks by evaluating the quality of 17 commonly used Turkish benchmark datasets. Using a comprehensive framework that assesses six criteria, both human and LLM-judge annotators provide detailed evaluations to identify dataset strengths and shortcomings. Our results reveal that 70% of the benchmark datasets fail to meet our heuristic quality standards. The correctness of the usage of technical terms is the strongest criterion, but 85% of the criteria are not satisfied in the examined datasets. Although LLM judges demonstrate potential, they are less effective than human annotators, particularly in understanding cultural common sense knowledge and interpreting fluent, unambiguous text. GPT-4o has stronger labeling capabilities for grammatical and technical tasks, while Llama3.3-70B excels at correctness and cultural knowledge evaluation. Our findings emphasize the urgent need for more rigorous quality control in creating and adapting datasets for low-resource languages.

We examine if frontier chat-based large language models (LLMs) adjust their outputs based on neurodivergence (ND) context in system prompts and describe the nature of these adjustments. Specifically, we propose NDBench, a 576-output benchmark involving two frontier models, three system prompt types (baseline, ND-profile assertion, and ND-profile assertion with explicit instructions for adjustments), four canonical ND profiles, and 24 prompts across four categories, one of which involves an adversarial masking strategy. Four trends emerge consistently from our findings. First, LLMs show significant adaptation under ND context, where fully instructed conditions yield lengthier and more structured outputs, characterized by higher token counts, more headings, and more granular steps (p < 10^-8, Holm-corrected). Second, such adaptation is largely structural in nature: although list density does not change much, there is a marked rise in the frequency of headings and per-step detail. Third, ND persona assertion alone fails to suppress potentially harmful tendencies, as masking-reinforcement decreases only in explicitly instructed cases (36-44% reduction); the reduction rate barely changes in persona assertion conditions. Moreover, reliability analysis of LLM-based harm assessment reveals that only two out of the six dimensions (masking and reinforcement, validation quality) exceed the pre-defined inter-judge agreement criterion (alpha >= 0.67) and thus can be considered primary results. NDBench is made publicly available along with its prompts, outputs, code, and other resources, forming a reproducible framework for auditing future LLMs' adaptation to ND awareness.

Large language models (LLMs) deployed for logical reasoning in knowledge-intensive domains exhibit a subtle but critical failure: coherence can be vacuously achieved through systematic abstention. A model that withholds commitment to either entailment or refutation satisfies negation consistency while providing no utility. We introduce Coherence Under Commitment (CUC), a dual-query evaluation paradigm that jointly measures consistency and decisiveness. CUC contributes three innovations: (1) a commitment score c(φ) = p(φ) + p(lnotφ) quantifying probability mass allocated to decisive outcomes; (2) a deterministic elicitation protocol via normalized YES/NO log probabilities, eliminating sampling variance; and (3) a 3-way decision framework (True/False/Uncertain) operationalizing the coherence-commitment trade-off into metrics. Experiments on four open-weight LLMs (1B-3B) across 204 FOLIO examples expose a sharp frontier. Qwen2.5-3B achieves near-zero contradiction (E[v_{neg}]{=}0.025) but only 7.4% coverage, while TinyLlama-1.1B reaches 79.4% coverage with violations on every example. Coherence-only evaluation would rank the abstaining model first; CUC exposes this as vacuous, and the frontier generalizes to LogiQA~v2 (ρ{=}0.97). We argue that evaluation must report both coherence and non-vacuous commitment and release a toolkit for standardized assessment.

The recent emergence of Large Vision-Language Models(VLMs) has resulted in a variety of different benchmarks for evaluating such models. Despite this, we observe that most existing evaluation methods suffer from the fact that they either require the model to choose from pre-determined responses, sacrificing open-endedness, or evaluate responses using a judge model, resulting in subjective and unreliable evaluation. In addition, we observe a lack of benchmarks for VLMs in the Korean language, which are necessary as a separate metric from more common English language benchmarks, as the performance of generative language models can differ significantly based on the language being used. Therefore, we present KOFFVQA, a general-purpose free-form visual question answering benchmark in the Korean language for the evaluation of VLMs. Our benchmark consists of 275 carefully crafted questions each paired with an image and grading criteria covering 10 different aspects of VLM performance. The grading criteria eliminate the problem of unreliability by allowing the judge model to grade each response based on a pre-determined set of rules. By defining the evaluation criteria in an objective manner, even a small open-source model can be used to evaluate models on our benchmark reliably. In addition to evaluating a large number of existing VLMs on our benchmark, we also experimentally verify that our method of using pre-existing grading criteria for evaluation is much more reliable than existing methods. Our evaluation code is available at https://github.com/maum-ai/KOFFVQA

Evaluation of multilingual Large Language Models (LLMs) is challenging due to a variety of factors -- the lack of benchmarks with sufficient linguistic diversity, contamination of popular benchmarks into LLM pre-training data and the lack of local, cultural nuances in translated benchmarks. In this work, we study human and LLM-based evaluation in a multilingual, multi-cultural setting. We evaluate 30 models across 10 Indic languages by conducting 90K human evaluations and 30K LLM-based evaluations and find that models such as GPT-4o and Llama-3 70B consistently perform best for most Indic languages. We build leaderboards for two evaluation settings - pairwise comparison and direct assessment and analyse the agreement between humans and LLMs. We find that humans and LLMs agree fairly well in the pairwise setting but the agreement drops for direct assessment evaluation especially for languages such as Bengali and Odia. We also check for various biases in human and LLM-based evaluation and find evidence of self-bias in the GPT-based evaluator. Our work presents a significant step towards scaling up multilingual evaluation of LLMs.

Indic quality estimation (QE) and automatic post-editing (APE) data is spread across separate releases, so no single resource supports training and evaluation across tasks and language pairs on one footing. We consolidate the WMT 2020--2024 shared-task lineage with an extended English--Malayalam resource into \indicqe: 126{,}754 instances over nine directional pairs, with up to four label types aligned on the same segment, a direct assessment, a human post-edit, word-level OK/BAD tags and an error explanation, and a test set stratified over four difficulty axes. On it, we benchmark six prompted LLMs and three COMET metrics on segment-level QE, and three systems on APE. Two of the axes are defined partly on the direct assessment and select a compressed slice of it, so each axis is compared against a control drawn from the same language pair with the same score distribution. Only one survives that control: segments whose holistic and token-level quality signals conflict are ranked worse than equally-scored segments of the same language, for all nine systems and all seven pairs that carry the axis. Annotator disagreement, which looks second-hardest without the control, has no effect with it. Few-shot prompting costs every model leq 3.4B both correlation and output-format compliance. Within-language accuracy does not make scores comparable across pairs: of the three trained metrics, the one with the best within-language correlation loses most when the pairs are pooled. The benchmark and code will be released.

Scalar reward models compress multi-dimensional human preferences into a single opaque score, creating an information bottleneck that often leads to brittleness and reward hacking in open-ended alignment. We argue that robust alignment for non-verifiable tasks is fundamentally a principle generalization problem: reward should not be a learned function internalized into a judge, but an explicit reasoning process executed under inspectable principles. To operationalize this view, we present the Open Rubric System (OpenRS), a plug-and-play, rubrics-based LLM-as-a-Judge framework built around Pairwise Adaptive Meta-Rubrics (PAMR) and lightweight Pointwise Verifiable Rubrics (PVRs), which provide both hard-constraint guardrails and verifiable reward components when ground-truth or programmatic checks are available. OpenRS uses an explicit meta-rubric -- a constitution-like specification that governs how rubrics are instantiated, weighted, and enforced -- and instantiates adaptive rubrics on the fly by conditioning on the semantic differences between two candidate responses. It then performs criterion-wise pairwise comparisons and aggregates criterion-level preferences externally, avoiding pointwise weighted scalarization while improving discriminability in open-ended settings. To keep principles consistent yet editable across various domains, we introduce a two-level meta-rubric refinement pipeline (automated evolutionary refinement for general principles and a reproducible human-in-the-loop procedure for domain principles), complemented with pointwise verifiable rubrics that act as both guardrails against degenerate behaviors and a source of verifiable reward for objective sub-tasks. Finally, we instantiate OpenRS as reward supervision in pairwise RL training.

Superalignment, where humans are weak supervisors of superhuman models, has become an important and widely discussed issue in the current era of rapid development of Large Language Models (LLMs). The recent work preliminarily studies this problem by using weak models to supervise strong models. It discovers that weakly supervised strong students can consistently outperform weak teachers towards the alignment target, leading to a weak-to-strong generalization phenomenon. However, we are concerned that behind such a promising phenomenon, whether there exists an issue of weak-to-strong deception, where strong models may deceive weak models by exhibiting well-aligned in areas known to weak models but producing misaligned behaviors in cases weak models do not know. We then take an initial step towards exploring this security issue in a specific but realistic multi-objective alignment case, where there may be some alignment targets conflicting with each other (e.g., helpfulness v.s. harmlessness). Such a conflict is likely to cause strong models to deceive weak models in one alignment dimension to gain high reward in other alignment dimension. Our experiments on both the reward modeling task and the preference optimization scenario indicate: (1) the weak-to-strong deception exists; (2) the deception phenomenon may intensify as the capability gap between weak and strong models increases. We also discuss potential solutions and find bootstrapping with an intermediate model can mitigate the deception to some extent. Our work highlights the urgent need to pay more attention to the true reliability of superalignment.

We introduce JP-TL-Bench, a lightweight, open benchmark designed to guide the iterative development of Japanese-English translation systems. In this context, the challenge is often "which of these two good translations is better?" rather than "is this translation acceptable?" This distinction matters for Japanese-English, where subtle choices in politeness, implicature, ellipsis, and register strongly affect perceived naturalness. JP-TL-Bench uses a protocol built to make LLM judging both reliable and affordable: it evaluates a candidate model via reference-free, pairwise LLM comparisons against a fixed, versioned anchor set. Pairwise results are aggregated with a Bradley-Terry model and reported as win rates plus a normalized 0-10 "LT" score derived from a logistic transform of fitted log-strengths. Because each candidate is scored against the same frozen anchor set, scores are structurally stable given the same base set, judge, and aggregation code.

Research planning is the decisive capability of AI scientists. Yet a research plan admits no verifiable answer, so reinforcement learning lacks the environment it requires: tasks paired with a critic. Rubrics extracted from scientific papers can supply the critic. Existing pipelines, however, draw the question and the criteria from the same content, so the reward can be earned by paraphrase. The rubric is further compressed into a single scalar per rollout. We introduce PaperGym, a unified framework that turns each research paper into a complete training environment. PaperGym exploits the structure of a paper: the question is synthesized from the research goal and background, while the criteria are derived from the method and experiments. The criteria span methodological innovation and experimental design, and criterion leakage falls to 3.7%, versus 11.90% to 34.10% in existing datasets. Training uses the rubric twice: first as privileged context for OPSD's self-teacher, then as the reward for GRPO. Across Qwen3-1.7B/4B/8B, this schedule outperforms supervised fine-tuning, either stage alone, and the reverse ordering, improving five-benchmark averages by +5.6, +5.0, and +4.8 points. With the recipe held fixed, models trained on PaperGym-20k win 58.1% of three-way comparisons, against 28.2% for RubricHub Science. The trained Qwen3-8B reaches 73.48 on ResearchQA, above the far larger Kimi K2.6. We release the pipeline, the 20,000-instance corpus PaperGym-20k, and the benchmarks PaperGym-Innov and PaperGym-Design.

Reliable application of machine learning is of primary importance to the practical deployment of deep learning methods. A fundamental challenge is that models are often unreliable due to overconfidence. In this paper, we estimate a model's reliability by measuring the agreement between its latent space, and the latent space of a foundation model. However, it is challenging to measure the agreement between two different latent spaces due to their incoherence, \eg, arbitrary rotations and different dimensionality. To overcome this incoherence issue, we design a neighborhood agreement measure between latent spaces and find that this agreement is surprisingly well-correlated with the reliability of a model's predictions. Further, we show that fusing neighborhood agreement into a model's predictive confidence in a post-hoc way significantly improves its reliability. Theoretical analysis and extensive experiments on failure detection across various datasets verify the effectiveness of our method on both in-distribution and out-of-distribution settings.

In legal NLP, Case Outcome Classification (COC) must not only be accurate but also trustworthy and explainable. Existing work in explainable COC has been limited to annotations by a single expert. However, it is well-known that lawyers may disagree in their assessment of case facts. We hence collect a novel dataset RAVE: Rationale Variation in ECHR1, which is obtained from two experts in the domain of international human rights law, for whom we observe weak agreement. We study their disagreements and build a two-level task-independent taxonomy, supplemented with COC-specific subcategories. To our knowledge, this is the first work in the legal NLP that focuses on human label variation. We quantitatively assess different taxonomy categories and find that disagreements mainly stem from underspecification of the legal context, which poses challenges given the typically limited granularity and noise in COC metadata. We further assess the explainablility of SOTA COC models on RAVE and observe limited agreement between models and experts. Overall, our case study reveals hitherto underappreciated complexities in creating benchmark datasets in legal NLP that revolve around identifying aspects of a case's facts supposedly relevant to its outcome.

Radiologists play a crucial role by translating medical images into medical reports. However, the field faces staffing shortages and increasing workloads. While automated approaches using vision-language models (VLMs) show promise as assistants, they require exceptionally high accuracy. Most current VLMs in radiology rely solely on supervised fine-tuning (SFT). Meanwhile, in the general domain, additional preference fine-tuning has become standard practice. The challenge in radiology lies in the prohibitive cost of obtaining radiologist feedback. We propose a scalable automated preference alignment technique for VLMs in radiology, focusing on chest X-ray (CXR) report generation. Our method leverages publicly available datasets with an LLM-as-a-Judge mechanism, eliminating the need for additional expert radiologist feedback. We evaluate and benchmark five direct alignment algorithms (DAAs). Our results show up to a 57.4% improvement in average GREEN scores, a LLM-based metric for evaluating CXR reports, and a 9.2% increase in an average across six metrics (domain specific and general), compared to the SFT baseline. We study reward overoptimization via length exploitation, with reports lengthening by up to 3.2x. To assess a potential alignment tax, we benchmark on six additional diverse tasks, finding no significant degradations. A reader study involving four board-certified radiologists indicates win rates of up to 0.62 over the SFT baseline, while significantly penalizing verbosity. Our analysis provides actionable insights for the development of VLMs in high-stakes fields like radiology.

The explosion of visual content available online underscores the requirement for an accurate machine assessor to robustly evaluate scores across diverse types of visual contents. While recent studies have demonstrated the exceptional potentials of large multi-modality models (LMMs) on a wide range of related fields, in this work, we explore how to teach them for visual rating aligned with human opinions. Observing that human raters only learn and judge discrete text-defined levels in subjective studies, we propose to emulate this subjective process and teach LMMs with text-defined rating levels instead of scores. The proposed Q-Align achieves state-of-the-art performance on image quality assessment (IQA), image aesthetic assessment (IAA), as well as video quality assessment (VQA) tasks under the original LMM structure. With the syllabus, we further unify the three tasks into one model, termed the OneAlign. In our experiments, we demonstrate the advantage of the discrete-level-based syllabus over direct-score-based variants for LMMs. Our code and the pre-trained weights are released at https://github.com/Q-Future/Q-Align.

Sycophancy in LLMs is documented across 70+ papers, but expert agreement on construct boundaries remains low (ICC=.184; Ye et al., 2026). The construct fragments because behavioral classification depends on which surface form is privileged. We adopt a materials-science framing: conversation as test specimen under load, LLM-model as material charge, pushback as progressive load, stance-flip as material failure. We characterize this failure across three loading cases (debate n=1000; false-presuppositions n=3400; ethical-setting n=3400; 10-17 material charges per case; 7800 specimens total) using 14 turn-level axis-measurements spanning velocity, damage accumulation, frame-drift, brittleness, and direction stability, plus three speaker-resolved axes from an independent pipeline. The measurements are Hooke-coupled (σ= E cdot varepsilon analog) and reproduce across loading cases with effects up to |r_{rb}| = 0.35 on debate; the sign structure adds a second pattern: the ethical-setting case inverts the velocity and accumulation blocks. Variance composition partitions into two profiles: debate is charge-dominated (brittle-fracture-like: the material grade decides), false-presuppositions and ethical-setting are topic-dominated (creep-like: the load decides); the ratios (2.03 vs 0.13/0.17) are estimator-dependent, for debate even in direction. Cross-judge reliability (GPT-4o vs Haiku 4.5) shows debate scoring is judge-robust (Cohen's κ= 0.88) while false-presupposition scoring is judge-sensitive (κ= 0.36) -- a caveat single-judge benchmarks must report. This is the methodological move Ye et al.'s diagnosis calls for: a multi-axis characterization that does not depend on which surface form of the construct one privileges.

Vision-language models (VLMs) often produce chain-of-thought (CoT) explanations that sound plausible yet fail to reflect the underlying decision process, undermining trust in high-stakes clinical use. Existing evaluations rarely catch this misalignment, prioritizing answer accuracy or adherence to formats. We present a clinically grounded framework for chest X-ray visual question answering (VQA) that probes CoT faithfulness via controlled text and image modifications across three axes: clinical fidelity, causal attribution, and confidence calibration. In a reader study (n=4), evaluator-radiologist correlations fall within the observed inter-radiologist range for all axes, with strong alignment for attribution (Kendall's tau_b=0.670), moderate alignment for fidelity (tau_b=0.387), and weak alignment for confidence tone (tau_b=0.091), which we report with caution. Benchmarking six VLMs shows that answer accuracy and explanation quality are decoupled, acknowledging injected cues does not ensure grounding, and text cues shift explanations more than visual cues. While some open-source models match final answer accuracy, proprietary models score higher on attribution (25.0% vs. 1.4%) and often on fidelity (36.1% vs. 31.7%), highlighting deployment risks and the need to evaluate beyond final answer accuracy.

The rapid advancement of Large Language Models (LLMs) necessitates robust and challenging benchmarks. Leaderboards like Chatbot Arena rank LLMs based on how well their responses align with human preferences. However, many tasks such as those related to emotional intelligence, creative writing, or persuasiveness, are highly subjective and often lack majoritarian human agreement. Judges may have irreconcilable disagreements about what constitutes a better response. To address the challenge of ranking LLMs on highly subjective tasks, we propose a novel benchmarking framework, the Language Model Council (LMC). The LMC operates through a democratic process to: 1) formulate a test set through equal participation, 2) administer the test among council members, and 3) evaluate responses as a collective jury. We deploy a council of 20 newest LLMs on an open-ended emotional intelligence task: responding to interpersonal dilemmas. Our results show that the LMC produces rankings that are more separable, robust, and less biased than those from any individual LLM judge, and is more consistent with a human-established leaderboard compared to other benchmarks.

This paper reports a single, fully instrumented case study of a large-scale architectural refactoring by an AI coding agent under a specification-first protocol, with no human review of the generated code and no pre-existing oracle to validate the target behaviour. The task, dismantling a central invariant across a large interdependent codebase, was assessed by the author as effectively infeasible through incremental refactoring, the kind of change that conventionally calls for a rewrite instead. Under the protocol described here, the agent completed it successfully. The system is a 717,725-line production TypeScript application across 3,648 files. The task required dismantling a core lifetime invariant: the guarantee that a UI panel remains open for the duration of an AI request. The target behaviour was that a streaming generation survives the closing of its panel and can be reattached, on reopening, to the same live stream with no loss or duplication. The protocol: formal specification by the agent, 14 refinement cycles auditing that specification against the source code, atomic implementation, a compile/test feedback loop, then 17 verification cycles auditing the code against the frozen specification. Across 31 audit passes, 201 defects were corrected before any human executed the program. The convergence criterion was empirical: two consecutive verification passes returning zero findings. The change touched 189 files (31 new); with the extraction phase, the two commits total 288 files, 34,770 insertions, 16,422 deletions. Across the first and roughly thirty later sessions, the software behaved as specified, no bug observed. Elapsed: three days; cost: USD 2,430. The full specification and raw session logs, 1,500+ pages in French, are published as evidence, allowing inspection of the process and submission to a language model for consistency checking.

Multimodal large language models (LLMs) are increasingly explored as automated evaluators in clinical settings, yet their scoring behavior on ordinal clinical scales remains poorly understood. We benchmark three frontier LLM families against supervised deep learning models for scoring Clock Drawing Test (CDT) images on two public datasets using the Shulman rubric. While fully fine-tuned Vision Transformers achieve the best calibration (MAE 0.52, within-1 accuracy 91%), zero-shot LLMs remain competitive on tolerance-based agreement (GPT-5 MAE 0.67, within-1 accuracy 92%) despite higher absolute error. However, per-score analysis reveals that all three LLM families exhibit a pronounced central tendency effect (systematic endpoint compression): predictions are systematically compressed toward the middle of the scale, with over-prediction at the low end (score 0 to 1) and under-prediction at the high end (score 5 to 4). This effect disproportionately affects the clinically critical extremes where accurate scoring most impacts screening decisions for cognitive impairment. Targeted ablations show that neither few-shot exemplars spanning the full score range nor removing clinical terminology from the prompt eliminates the effect. Our findings extend the LLM-as-a-judge bias literature from NLP evaluation to clinical assessment, and highlight the need for calibration-aware evaluation and post-hoc calibration before deploying LLM-based raters in high-stakes screening workflows.

Autonomous research agents produce competitive solutions and professional-looking manuscripts, yet their outputs contain verifiability failures undetectable by surface-level evaluation: fabricated citations, unreproducible scores, and method descriptions that diverge from the implementation. We address this through three contributions. First, Chain-of-Evidence (CoE), a verifiability framework requiring every claim to be traceable to its evidence source. Second, ScientistOne, an end-to-end autonomous research system that maintains evidence chains by construction throughout literature review, solution discovery, and paper writing. Third, CoE Audit, a post-hoc audit whose four integrity checks -- score verification, specification violation, reference verification, and method-code alignment -- apply uniformly to all systems. Across 75 papers spanning five systems and five frontier research tasks, every baseline exhibits at least one systematic failure mode: hallucinated reference rates reach 21%, score verification passes in as few as 42% of papers, and method-code alignment ranges from 20% to 80%. ScientistOne achieves zero hallucinated references (0/337), perfect score verification (12/12), and the highest method-code alignment (14/15), while matching or exceeding human expert performance on all five tasks. ScientistOne further generalizes to six additional tasks spanning medical imaging, fine-grained recognition, 3D perception, and language modeling, achieving state-of-the-art on Parameter Golf and gold medals on MLE-Bench tasks where baselines fail entirely.

LLM-generated reviews for scientific papers are gaining considerable traction and are even being officially piloted by major conferences. We have to assume that not only reviewers are using LLM-assistance, but also that authors use LLMs to revise their papers before submitting. In this work, we perform empirical experiments on papers from the 2025 ACL Rolling Review (ARR) to evaluate LLM reviews from both the author and the reviewer perspective. First, we identify a limited alignment of LLM reviews with human ones. In the best-case scenario, the alignment is reasonable. However, we also find that LLM-human alignment varies substantially across prompts and models. Finally, we investigate the scenario in which the author uses an iterative draft-revise workflow to improve the submission according to the LLM review. We find that this "gaming" of LLM reviews can be effective in specific scenarios, leading to a statistically significant increase of overall scores for up to 35\% of papers. We publish our code: https://github.com/uhh-hcds/reviewarcade.

LLM-as-a-Judge has been widely adopted as an evaluation method and served as supervised rewards in model training. However, existing benchmarks for LLM-as-a-Judge are mainly relying on human-annotated ground truth, which introduces human bias that undermines the assessment of reliability and imposes scalability constraints. To overcome these limitations, we introduce Sage, a novel evaluation suite that assesses the quality of LLM judges without necessitating any human annotation. Inspired by axioms of rational choice theory, Sage introduces two new lenses for measuring LLM-as-a-Judge: local self-consistency (pair-wise preference stability) and global logical consistency (transitivity across a full set of preferences). We curate a dataset of 650 questions by combining structured benchmark problems with real-world user queries. Our experiments demonstrate both the stability of our metrics and their high correlation with supervised benchmarks like LLMBar and RewardBench2, confirming Sage's reliability as an evaluation suite for the robustness and accuracy of LLM-as-a-Judge. Based on Sage, we reveal that current state-of-the-art LLMs exhibit significant reliability problems when acting as judges in both scoring and pairwise settings; even the top-performing models, Gemini-2.5-Pro and GPT-5, fail to maintain consistent preferences in nearly a quarter of difficult cases. We attribute this to a new phenomenon called situational preference, which explains why explicit rubrics or criteria can help the model judge consistently across answer pairs. Our further analysis shows that finetuned LLM-as-a-Judge is a feasible method to boost performance, and the panel-based judge as well as deep reasoning can enhance the judging consistency. We also find substantial inconsistency in human judgments, which indicates that human annotation may not be a reliable gold standard.

LLM-as-a-Judge is a framework that uses an LLM (large language model) to evaluate the quality of natural language text - typically text that is also generated by an LLM. This framework holds great promise due to its relative low-cost, ease of use, and strong correlations with human stylistic preferences. However, LLM Judges have been shown to exhibit biases that can distort their judgments. We evaluate how well LLM Judges can grade whether a given response to a conversational question is correct, an ability crucial to soundly estimating the overall response quality. To do so, we create and publicly release a human-annotated dataset with labels of correctness for 1,200 LLM responses. We source questions from a combination of existing datasets and a novel, challenging benchmark (BFF-Bench) created for this analysis. We demonstrate a strong connection between an LLM's ability to correctly answer a question and grade responses to that question. Although aggregate level statistics might imply a judge has high agreement with human annotators, it will struggle on the subset of questions it could not answer. To address this issue, we recommend a simple solution: provide the judge with a correct, human-written reference answer. We perform an in-depth analysis on how reference quality can affect the performance of an LLM Judge. We show that providing a weaker judge (e.g. Qwen 2.5 7B) with higher quality references reaches better agreement with human annotators than a stronger judge (e.g. GPT-4o) with synthetic references.

Evaluation of language model outputs on structured writing tasks is typically conducted with a number of desirable criteria presented to human evaluators or large language models (LLMs). For instance, on a prompt like "Help me draft an academic talk on coffee intake vs research productivity", a model response may be evaluated for criteria like accuracy and coherence. However, high-quality responses should do more than just satisfy basic task requirements. An effective response to this query should include quintessential features of an academic talk, such as a compelling opening, clear research questions, and a takeaway. To help identify these implicit criteria, we introduce EvalAgent, a novel framework designed to automatically uncover nuanced and task-specific criteria. EvalAgent first mines expert-authored online guidance. It then uses this evidence to propose diverse, long-tail evaluation criteria that are grounded in reliable external sources. Our experiments demonstrate that the grounded criteria produced by EvalAgent are often implicit (not directly stated in the user's prompt), yet specific (high degree of lexical precision). Further, EvalAgent criteria are often not satisfied by initial responses but they are actionable, such that responses can be refined to satisfy them. Finally, we show that combining LLM-generated and EvalAgent criteria uncovers more human-valued criteria than using LLMs alone.

Urban assessments often compress diverse needs into single scores, which can obscure minority perspectives. We present a community-centered study in Montreal (n=35; wheelchair users, seniors, LGBTQIA2+ residents, and immigrants). Participants rated 20 streets (accessibility, inclusivity, aesthetics, practicality) and ranked 7 images on 12 interview-elicited criteria. Disagreement patterns were systematic in our sample: wheelchair users diverged most on accessibility and practicality; LGBTQIA2+ participants emphasized inclusion and liveliness; seniors prioritized security. Group discussion reduced information gaps but not value conflicts; ratings conveyed intensity, while rankings forced trade-offs. We then formalize negotiative alignment, a transparent, budget-aware bargaining procedure, and pilot it with role-played stakeholder agents plus a neutral mediator. Relative to the best base design under the same public rubric, the negotiated package increased total utility (21.10 to 24.55), raised the worst-group utility (3.20 to 3.90), improved twentieth percentile satisfaction (0.86 to 1.00; min-max normalized within the scenario), and reduced inequality (Gini 0.036 to 0.025). Treating disagreement as signal and reporting worst-group outcomes alongside totals may help planners and AI practitioners surface trade-offs and preserve minority priorities while maintaining efficiency.

In medicine, claims remain valid when supported by empirical evidence grounded in stable biological reality. In law, by contrast, truth is contingent, defined by jurisdiction, temporal validity, and the hierarchy of authoritative sources. The recent success of large language models (LLMs) on medical licensing examinations has encouraged an expectation of comparable legal competence. This analogy, however, obscures a critical distinction between domains. Unlike in medicine, legal performance often depends less on inference than on determining when external authority is applicable, valid, and non-contradictory. We introduce a comparative diagnostic framework evaluating legal reasoning against medical baselines along four axes (knowledge recall, grounding, confidence, and robustness), uncovering a sharp domain asymmetry when applied to a new benchmark that encodes temporal validity and normative relationships. While medical LLMs reliably benefit from verified sources, legal LLMs struggle to assess when retrieved citations are useful or misleading, exhibiting overconfidence in perturbed contexts and sensitivity to superficial formatting cues. Increased model scale amplifies this tendency, revealing that stronger instruction following can coincide with weaker resistance to authoritative perturbations. These findings show that LLMs treat law as unstructured text rather than binding precedent, while revealing a tendency to over-trust authoritative but false information when external references conflict with a model's internal knowledge.

Recent work on chain-of-thought (CoT) faithfulness reports single aggregate numbers (e.g., DeepSeek-R1 acknowledges hints 39% of the time), implying that faithfulness is an objective, measurable property of a model. This paper demonstrates that it is not. Three classifiers (a regex-only detector, a two-stage regex-plus-LLM pipeline, and an independent Claude Sonnet 4 judge) are applied to 10,276 influenced reasoning traces from 12 open-weight models spanning 9 families and 7B to 1T parameters. On identical data, these classifiers produce overall faithfulness rates of 74.4%, 82.6%, and 69.7%, respectively, with non-overlapping 95% confidence intervals. Per-model gaps range from 2.6 to 30.6 percentage points; all are statistically significant (McNemar's test, p < 0.001). The disagreements are systematic, not random: inter-classifier agreement measured by Cohen's kappa ranges from 0.06 ("slight") for sycophancy hints to 0.42 ("moderate") for grader hints, and the asymmetry is pronounced: for sycophancy, 883 cases are classified as faithful by the pipeline but unfaithful by the Sonnet judge, while only 2 go the other direction. Classifier choice can also reverse model rankings: Qwen3.5-27B ranks 1st under the pipeline but 7th under the Sonnet judge; OLMo-3.1-32B moves in the opposite direction, from 9th to 3rd. The root cause is that different classifiers operationalize related faithfulness constructs at different levels of stringency (lexical mention versus epistemic dependence), and these constructs yield divergent measurements on the same behavior. These results demonstrate that published faithfulness numbers cannot be meaningfully compared across studies that use different classifiers, and that future evaluations should report sensitivity ranges across multiple classification methodologies rather than single point estimates.

Large language models (LLMs) increasingly influence global digital ecosystems, yet their potential to perpetuate social and cultural biases remains poorly understood in underrepresented contexts. This study presents a systematic analysis of representational biases in seven state-of-the-art LLMs: GPT-4o-mini, Claude-3-Sonnet, Claude-4-Sonnet, Gemini-2.0-Flash, Gemini-2.0-Lite, Llama-3-70B, and Mistral-Nemo in the Nepali cultural context. Using Croissant-compliant dataset of 2400+ stereotypical and anti-stereotypical sentence pairs on gender roles across social domains, we implement an evaluation framework, Dual-Metric Bias Assessment (DMBA), combining two metrics: (1) agreement with biased statements and (2) stereotypical completion tendencies. Results show models exhibit measurable explicit agreement bias, with mean bias agreement ranging from 0.36 to 0.43 across decoding configurations, and an implicit completion bias rate of 0.740-0.755. Importantly, implicit completion bias follows a non-linear, U-shaped relationship with temperature, peaking at moderate stochasticity (T=0.3) and declining slightly at higher temperatures. Correlation analysis under different decoding settings revealed that explicit agreement strongly aligns with stereotypical sentence agreement but is a weak and often negative predictor of implicit completion bias, indicating generative bias is poorly captured by agreement metrics. Sensitivity analysis shows increasing top-p amplifies explicit bias, while implicit generative bias remains largely stable. Domain-level analysis shows implicit bias is strongest for race and sociocultural stereotypes, while explicit agreement bias is similar across gender and sociocultural categories, with race showing the lowest explicit agreement. These findings highlight the need for culturally grounded datasets and debiasing strategies for LLMs in underrepresented societies.

This paper calls on the research community not only to investigate how human biases are inherited by large language models (LLMs) but also to explore how these biases in LLMs can be leveraged to make society's "unwritten code" - such as implicit stereotypes and heuristics - visible and accessible for critique. We introduce a conceptual framework through a case study in science: uncovering hidden rules in peer review - the factors that reviewers care about but rarely state explicitly due to normative scientific expectations. The idea of the framework is to push LLMs to speak out their heuristics through generating self-consistent hypotheses - why one paper appeared stronger in reviewer scoring - among paired papers submitted to 45 computer science conferences, while iteratively searching deeper hypotheses from remaining pairs where existing hypotheses cannot explain. We observed that LLMs' normative priors about the internal characteristics of good science extracted from their self-talk, e.g. theoretical rigor, were systematically updated toward posteriors that emphasize storytelling about external connections, such as how the work is positioned and connected within and across literatures. This shift reveals the primacy of scientific myths about intrinsic properties driving scientific excellence rather than extrinsic contextualization and storytelling that influence conceptions of relevance and significance. Human reviewers tend to explicitly reward aspects that moderately align with LLMs' normative priors (correlation = 0.49) but avoid articulating contextualization and storytelling posteriors in their review comments (correlation = -0.14), despite giving implicit reward to them with positive scores. We discuss the broad applicability of the framework, leveraging LLMs as diagnostic tools to surface the tacit codes underlying human society, enabling more precisely targeted responsible AI.

Evaluating large language models (LLMs) on open-ended questions is difficult because response quality depends on the question's context. Binary scores and static rubrics fail to capture these context-dependent requirements. Existing methods define criteria at the dataset level or generate them in a single pass, which limits their ability to explore the evaluation space implied by each question. We introduce One-Question-One-World (Qworld), a method that generates question-specific evaluation criteria using a recursive expansion tree. Given a question, Qworld decomposes it into scenarios, perspectives, and fine-grained binary criteria through structured hierarchical and horizontal expansion. The resulting criteria specify what a high-quality answer must address for that question. On HealthBench, Qworld covers 89% of expert-authored criteria and generates 79% novel criteria validated by human experts. Experts rate Qworld criteria higher in insight and granularity than those produced by prior methods. When applied to 11 frontier LLMs on HealthBench and Humanity's Last Exam, Qworld reveals capability differences in dimensions such as long-term impact, equity, error handling, and interdisciplinary reasoning that coarse rubrics do not distinguish. By formulating criteria generation as structured coverage of question-implied evaluation axes, Qworld enables evaluation that adapts to each question rather than relying on fixed task-level criteria.

We introduce the first version of KWBench (Knowledge Work Bench), a benchmark for unprompted problem recognition in large language models: can an LLM identify a professional scenario before attempting to solve it. Existing frontier benchmarks have saturated, and most knowledge-work evaluations to date reduce to extraction or task completion against a specification. KWBench targets the step before that: recognizing the governing structure of the situation from raw inputs alone. The benchmark contains 223 tasks sourced from practitioners across acquisitions, contract negotiations, clinical pharmacy, organizational politics, fraud analysis, and incentive design. Each task encodes a formal game-theoretic pattern (principal-agent conflict, signaling, mechanism design failure, strategic omission, coalitional dynamics, strategic interdependence) and carries structured ground truth recording the expert reading of the situation and the anticipated failure modes. Models receive raw data and a task prompt with no indication of problem type. Scoring is a three-tier rubric gated by a mandatory conjunctive check. Mandatory criteria encode the predicted wrong paths. We evaluate 16 models. The best model passes on 27.9% of tasks. The top two models agree on only 31.7% of their passes. Among the top 8, 44 tasks are solved by exactly one model; routing across the top 8 covers 50.7% of the benchmark, nearly double the best single model. Conditional on passing, quality scores converge (approx 83% across models); unconditional scores do not. Same models articulate the relevant game-theoretic concept correctly when asked, then fail to apply it unprompted. We release KWBench to shift how frontier models are evaluated on knowledge work, scoring them on whether they recognize the right problem from the situation alone, not only on how well they execute once the problem has been framed for them.

We introduce the Agent GPA (Goal-Plan-Action) framework: an evaluation paradigm based on an agent's operational loop of setting goals, devising plans, and executing actions. The framework includes five evaluation metrics: Goal Fulfillment, Logical Consistency, Execution Efficiency, Plan Quality, and Plan Adherence. Logical Consistency checks that an agent's actions are consistent with its prior actions. Execution Efficiency checks whether the agent executes in the most efficient way to achieve its goal. Plan Quality checks whether an agent's plans are aligned with its goals; Plan Adherence checks if an agent's actions are aligned with its plan; and Goal Fulfillment checks that agent's final outcomes match the stated goals. Our experimental results on two benchmark datasets - the public TRAIL/GAIA dataset and an internal dataset for a production-grade data agent - show that this framework (a) provides a systematic way to cover a broad range of agent failures, including all agent errors on the TRAIL/GAIA benchmark dataset; (b) supports LLM-judges that exhibit strong agreement with human annotation, covering 80% to over 95% errors; and (c) localizes errors with 86% agreement to enable targeted improvement of agent performance.

Behaviour-Driven Development (BDD) suites accumulate step-text duplication whose maintenance cost is established in prior work. Existing detection techniques require running the tests (Binamungu et al., 2018-2023) or are confined to a single organisation (Irshad et al., 2020-2022), leaving a gap: a purely static, paraphrase-robust, step-level detector usable on any repository. We fill the gap with cukereuse, an open-source Python CLI combining exact hashing, Levenshtein ratio, and sentence-transformer embeddings in a layered pipeline, released alongside an empirical corpus of 347 public GitHub repositories, 23,667 parsed .feature files, and 1,113,616 Gherkin steps. The step-weighted exact-duplicate rate is 80.2 %; the median-repository rate is 58.6 % (Spearman rho = 0.51 with size). The top hybrid cluster groups 20.7k occurrences across 2.2k files. Against 1,020 pairs manually labelled by the three authors under a released rubric (inter-annotator Fleiss' kappa = 0.84 on a 60-pair overlap), we report precision, recall, and F1 with bootstrap 95 % CIs under two protocols: the primary rubric and a score-free second-pass relabelling. The strongest honest pair-level number is near-exact at F1 = 0.822 on score-free labels; the primary-rubric semantic F1 = 0.906 is inflated by a stratification artefact that pins recall at 1.000. Lexical baselines (SourcererCC-style, NiCad-style) reach primary F1 = 0.761 and 0.799. The paper also presents a CDN-structured critique of Gherkin (Cognitive Dimensions of Notations); eight of fourteen dimensions are rated problematic or unsupported. The tool, corpus, labelled pairs, rubric, and pipeline are released under permissive licences.

Large language models (LLMs) often hallucinate by generating factually incorrect or unfaithful content, posing significant risks to their safe use. Detecting such hallucinations is particularly challenging under the zero-source constraint, where no model internals or external references are available, and detection must rely solely on the textual query-answer pair. In this paper, we propose Human-like Criteria Probing for Hallucination Detection (HCPD), a paradigm that emulates the multi-faceted reasoning of human evaluators. Its core is a Human-like Criteria Probing (HCP) mechanism, in which a LLM agent adaptively decomposes its judgment into a weighted set of interpretable criteria and aggregates criterion-specific scores into a final truthfulness measure. To achieve this adaptive capability, we introduce a reward-based alignment scheme using only weak supervision from semantic consistency. At inference, we employ a multi-sampling aggregation strategy to ensure robust decisions while preserving full interpretability. We further provide theoretical analysis supporting the reliability of our approach. Extensive experiments show that HCPD consistently outperforms state-of-the-art baselines, offering an effective and explainable solution for zero-source hallucination detection. Code is available at https://github.com/TRISKEL10N/HCPD.

Despite notable advancements in Multimodal Large Language Models (MLLMs), most state-of-the-art models have not undergone thorough alignment with human preferences. This gap exists because current alignment research has primarily achieved progress in specific areas (e.g., hallucination reduction), while the broader question of whether aligning models with human preferences can systematically enhance MLLM capability remains largely unexplored. To this end, we introduce MM-RLHF, a dataset containing 120k fine-grained, human-annotated preference comparison pairs. This dataset represents a substantial advancement over existing resources, offering superior size, diversity, annotation granularity, and quality. Leveraging this dataset, we propose several key innovations to improve both the quality of reward models and the efficiency of alignment algorithms. Notably, we introduce a Critique-Based Reward Model, which generates critiques of model outputs before assigning scores, offering enhanced interpretability and more informative feedback compared to traditional scalar reward mechanisms. Additionally, we propose Dynamic Reward Scaling, a method that adjusts the loss weight of each sample according to the reward signal, thereby optimizing the use of high-quality comparison pairs. Our approach is rigorously evaluated across 10 distinct dimensions and 27 benchmarks, with results demonstrating significant and consistent improvements in model performance. Specifically, fine-tuning LLaVA-ov-7B with MM-RLHF and our alignment algorithm leads to a 19.5% increase in conversational abilities and a 60% improvement in safety. We have open-sourced the preference dataset, reward model, training and evaluation code, as well as reward modeling and safety benchmarks. For more details, please visit our project page: https://mm-rlhf.github.io.

Multimodal Large Language Models (MLLMs) have significantly advanced document understanding, yet current Doc-VQA evaluations score only the final answer and leave the supporting evidence unchecked. This answer-only approach masks a critical failure mode: a model can land on the correct answer while grounding it in the wrong passage -- a critical risk in high-stakes domains like law, finance, and medicine, where every conclusion must be traceable to a specific source region. To address this, we introduce CiteVQA, a benchmark that requires models to return element-level bounding-box citations alongside each answer, evaluating both jointly. CiteVQA comprises 1,897 questions across 711 PDFs spanning seven domains and two languages, averaging 40.6 pages per document. To ensure fidelity and scalability, the ground-truth citations are generated by an automated pipeline-which identifies crucial evidence via masking ablation-and are subsequently validated through expert review. At the core of our evaluation is Strict Attributed Accuracy (SAA), which credits a prediction only when the answer and the cited region are both correct. Auditing 20 MLLMs reveals a pervasive Attribution Hallucination: models frequently produce the right answer while citing the wrong region. The strongest system (Gemini-3.1-Pro-Preview) achieves an SAA of only 76.0, and the strongest open-source MLLM reaches just 22.5. Ultimately, towards trustworthy document intelligence, CiteVQA exposes a reliability gap that answer-only evaluations overlook, providing the instrumentation needed to close it. Our repository is available at https://github.com/opendatalab/CiteVQA.

Advances in training, post-training, and inference-time methods have enabled frontier reasoning models to win gold medals in math competitions and settle challenging open problems. Gaining trust in the responses of these models requires that natural language proofs be checked for errors. LLM judges are increasingly being adopted to meet the growing demand for evaluating such proofs. While verification is considered easier than generation, what model capability does reliable verification actually require? We systematically evaluate four open-source and two frontier LLMs on datasets of human-graded natural language proofs of competition-level problems. We consider two key metrics: verifier accuracy and self-consistency (the rate of agreement across repeated judgments on the same proof). We observe that smaller open-source models are only up to ~10% behind frontier models in accuracy but they are up to ~25% more inconsistent. Furthermore, we see that verifier accuracy is sensitive to prompt choice across all models. We then demonstrate that the smaller models, in fact, do possess the mathematical capabilities to verify proofs at the level of frontier models, but they struggle to reliably elicit these capabilities with general judging prompts. Through an LLM-guided prompt search, we synthesize an ensemble of specialized prompts that overcome the specific failure modes of smaller models, boosting their performance by up to 9.1% in accuracy and 15.9% in self-consistency. These gains are realized across models and datasets, allowing models like Qwen3.5-35B to perform on par with frontier models such as Gemini 3.1 Pro for proof verification.

While large language models (LLMs) achieve near-perfect scores on medical licensing exams, these evaluations inadequately reflect the complexity and diversity of real-world clinical practice. We introduce MedHELM, an extensible evaluation framework for assessing LLM performance for medical tasks with three key contributions. First, a clinician-validated taxonomy spanning 5 categories, 22 subcategories, and 121 tasks developed with 29 clinicians. Second, a comprehensive benchmark suite comprising 35 benchmarks (17 existing, 18 newly formulated) providing complete coverage of all categories and subcategories in the taxonomy. Third, a systematic comparison of LLMs with improved evaluation methods (using an LLM-jury) and a cost-performance analysis. Evaluation of 9 frontier LLMs, using the 35 benchmarks, revealed significant performance variation. Advanced reasoning models (DeepSeek R1: 66% win-rate; o3-mini: 64% win-rate) demonstrated superior performance, though Claude 3.5 Sonnet achieved comparable results at 40% lower estimated computational cost. On a normalized accuracy scale (0-1), most models performed strongly in Clinical Note Generation (0.73-0.85) and Patient Communication & Education (0.78-0.83), moderately in Medical Research Assistance (0.65-0.75), and generally lower in Clinical Decision Support (0.56-0.72) and Administration & Workflow (0.53-0.63). Our LLM-jury evaluation method achieved good agreement with clinician ratings (ICC = 0.47), surpassing both average clinician-clinician agreement (ICC = 0.43) and automated baselines including ROUGE-L (0.36) and BERTScore-F1 (0.44). Claude 3.5 Sonnet achieved comparable performance to top models at lower estimated cost. These findings highlight the importance of real-world, task-specific evaluation for medical use of LLMs and provides an open source framework to enable this.

Reliable reward models are essential for text-to-video evaluation and alignment. However, the trade-off between evaluation accuracy and inference efficiency places high demands on the quality of training supervision. Existing approaches often rely on holistic judges with fixed rubrics or open-ended reasoning, leading to incomplete inspection, unfaithful justification, and entangled attribution. We introduce FIRM-Video, a unified checklist-driven data construction framework based on a check-before-score principle: construct dimension-specific checklists, verify each criterion against temporal visual evidence, and aggregate only verified decisions. For Instruction Following, FIRM-Video decomposes prompts into weighted atomic requirements; for World Coherence, it constructs prompt-calibrated, target-specific checks grounded in visible entities and actions; and for Perceptual Quality, it applies a generic taxonomy of visual defects. The verified criteria and scores are further transformed into natural-language analyses for end-to-end reward modeling. Subsequently, we construct FIRM-Video-90K with 88,044 dimension-specific instances from 29,348 videos, and introduce FIRM-Video-Bench with 750 point-wise human annotations across 250 videos. The Qwen3-VL-based FIRM-Video-8B achieves the best overall MAE on FIRM-Video-Bench while consistently delivering the highest VBench Total, Quality, and Semantic Scores in Best-of-8 sampling across three video generators.

In Machine Translation (MT) evaluation, metric performance is assessed based on agreement with human judgments. In recent years, automatic metrics have demonstrated increasingly high levels of agreement with humans. To gain a clearer understanding of metric performance and establish an upper bound, we incorporate human baselines in the MT meta-evaluation, that is, the assessment of MT metrics' capabilities. Our results show that human annotators are not consistently superior to automatic metrics, with state-of-the-art metrics often ranking on par with or higher than human baselines. Despite these findings suggesting human parity, we discuss several reasons for caution. Finally, we explore the broader implications of our results for the research field, asking: Can we still reliably measure improvements in MT evaluation? With this work, we aim to shed light on the limits of our ability to measure progress in the field, fostering discussion on an issue that we believe is crucial to the entire MT evaluation community.

Lectures are a learning experience for both students and teachers. Students learn from teachers about the subject material, while teachers learn from students about how to refine their instruction. However, online student feedback is unstructured and abundant, making it challenging for teachers to learn and improve. We take a step towards tackling this challenge. First, we contribute a dataset for studying this problem: SIGHT is a large dataset of 288 math lecture transcripts and 15,784 comments collected from the Massachusetts Institute of Technology OpenCourseWare (MIT OCW) YouTube channel. Second, we develop a rubric for categorizing feedback types using qualitative analysis. Qualitative analysis methods are powerful in uncovering domain-specific insights, however they are costly to apply to large data sources. To overcome this challenge, we propose a set of best practices for using large language models (LLMs) to cheaply classify the comments at scale. We observe a striking correlation between the model's and humans' annotation: Categories with consistent human annotations (>0.9 inter-rater reliability, IRR) also display higher human-model agreement (>0.7), while categories with less consistent human annotations (0.7-0.8 IRR) correspondingly demonstrate lower human-model agreement (0.3-0.5). These techniques uncover useful student feedback from thousands of comments, costing around 0.002$ per comment. We conclude by discussing exciting future directions on using online student feedback and improving automated annotation techniques for qualitative research.

Current benchmarks for AI clinician systems, often based on multiple-choice exams or manual rubrics, fail to capture the depth, robustness, and safety required for real-world clinical practice. To address this, we introduce the GAPS framework, a multidimensional paradigm for evaluating Grounding (cognitive depth), Adequacy (answer completeness), Perturbation (robustness), and Safety. Critically, we developed a fully automated, guideline-anchored pipeline to construct a GAPS-aligned benchmark end-to-end, overcoming the scalability and subjectivity limitations of prior work. Our pipeline assembles an evidence neighborhood, creates dual graph and tree representations, and automatically generates questions across G-levels. Rubrics are synthesized by a DeepResearch agent that mimics GRADE-consistent, PICO-driven evidence review in a ReAct loop. Scoring is performed by an ensemble of large language model (LLM) judges. Validation confirmed our automated questions are high-quality and align with clinician judgment. Evaluating state-of-the-art models on the benchmark revealed key failure modes: performance degrades sharply with increased reasoning depth (G-axis), models struggle with answer completeness (A-axis), and they are highly vulnerable to adversarial perturbations (P-axis) as well as certain safety issues (S-axis). This automated, clinically-grounded approach provides a reproducible and scalable method for rigorously evaluating AI clinician systems and guiding their development toward safer, more reliable clinical practice.

State machine replication (SMR) and Byzantine fault-tolerant (BFT) consensus guarantee agreement despite a bounded number of arbitrary, colluding faulty participants. However, these guarantees rely on participants outside this set correctly executing the protocol's transition semantics. Agentic validators expose a weaker boundary: an authenticated, responsive, non-equivocating, and protocol-compliant reasoning participant may still endorse a semantically invalid transition due to reasoning errors. We call this failure mode an epistemic fault, and the collective phenomenon the Honest Quorum Problem (where "honest" means protocol-compliant, not semantically correct). Such a quorum can satisfy ordinary checks while forming a certificate for an invalid transition. Thus, agreement alone does not guarantee semantic validity or execution safety. Furthermore, because agentic validators often share model weights, training distributions, prompts, or toolchains, they are highly susceptible to correlated epistemic faults. We define Epistemic Byzantine Fault Tolerance (EBFT), a fault-tolerance model for agentic infrastructure and post-deterministic distributed systems. EBFT augments the conventional Byzantine fault bound with two separate, confidence-indexed quantities: e_δ bounds coherent invalid endorsements outside the Byzantine set, and u_ε bounds unusable validator support that degrades liveness. These quantities characterize semantic safety risk and liveness degradation independently. We derive quorum-threshold conditions for semantic validity, consensus agreement, liveness, and feasible threshold selection, and outline a calibration methodology for estimating these budgets. We show that adding nominally distinct agents improves fault tolerance only when it measurably reduces the upper-tail concentration of invalid endorsements or unusable support.

Existing evaluations often reduce agent failures to system-level outcomes, obscuring where the fault originated and which intervention would improve the agent system. This creates a repair-assignment problem: the same visible failure may call for model post-training, harness engineering, environment redesign, or benchmark repair depending on its source. Because agent behavior emerges from interactions among models, harnesses, users, tools, memory, and environments, outcome-level labels are often insufficient for improvement. Most failure taxonomies do little to resolve this problem because they are benchmark-specific and lack a shared structure. We introduce an interaction-centric taxonomy that localizes failures to the interactions in which they originate and identifies the responsible component. It organizes 41 failure modes by assigning each to an edge between two components and a fault side indicating where the repair belongs. This makes the taxonomy actionable: model-side failures identify targets for post-training, harness-side failures point to scaffolding and tool-integration fixes, and environment or grader failures reveal evaluation conditions requiring redesign. The schema applies across agent architectures, from coding assistants to long-horizon personal assistants and multi-agent systems. We ground the taxonomy in worked examples from public benchmarks, model system cards, published reports, and logged agent trajectories, and evaluate its reproducibility using independent reasoning agents as judges. Across four frontier models, the strongest judge reaches Cohen's κ=0.76 against human category labels, suggesting that the categories capture shared structure rather than annotator-specific preferences.

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.

LLM confidence signals are used for abstention, routing, and safety-critical decisions. No standard practice exists for checking whether a confidence signal carries item-level information before building on it. We transfer the validity screening principle from clinical personality assessment (PAI, MMPI-3) as a portable protocol for benchmark-based LLM confidence data. The protocol specifies three core indices (L, Fp, RBS), a structural indicator (TRIN), and an item-sensitivity statistic, computed from a single 2x2 contingency table. A three-tier classification system (Invalid, Indeterminate, Valid) draws on four clinical traditions. Validated on 20 frontier LLMs across 524 items, four models are classified Invalid, two Indeterminate. Valid-profile models show mean r = .18 (15/16 significant). Invalid-profile models show mean r = -.20 (d = 2.48). Cross-benchmark validation on 18 models using MMLU with verbalized confidence and on external data from Yang et al. (2024) confirms the screen transfers across benchmarks and probe formats. All data and code: https://github.com/synthiumjp/validity-scaling-llm

As large language models enter professional domains, they must satisfy domain constraints, include critical evidence, and provide complete reasoning rather than merely produce fluent responses. Existing post-training methods often rely on holistic preferences or outcome-level verification, while recent rubric-based methods usually generate rubrics independently for each query. In specialized domains, such unconstrained rubrics may omit critical requirements and vary across samples, hindering the diagnosis and targeted repair of persistent capability deficiencies. We propose APTER (Adaptive Post-Training with Expert-Grounded Rubrics), a framework that integrates structured domain knowledge into fine-grained evaluation, optimization, and diagnosis for specialized complex reasoning. First, expert-grounded rubric construction starts from an expert criteria framework built by domain experts, where each criterion represents a stable professional capability. For each query, APTER selects relevant criteria and instantiates them into query-level rubrics linked to their source criteria, turning reusable expert criteria into executable query-level supervision without reference answers. Second, adaptive post-training uses rubric verdicts as both optimization and criterion-level diagnostic signals. Aggregating low-scoring verdicts by criterion ID reveals persistent deficiencies and triggers targeted supervised fine-tuning updates during reinforcement learning. Experiments on mathematical reasoning and medical question answering show consistent gains across both domains. Across three model generations, APTER improves the mathematics and medical averages over the corresponding base models by up to 15.86 and 8.04 points, respectively. Code and rubric datasets are available at https://github.com/AntDT-APTER/APTER.

Automated LLM-as-a-Judge frameworks have become the de facto standard for scalable evaluation across natural language processing. For instance, in safety evaluation, these judges are relied upon to evaluate harmfulness in order to benchmark the robustness of safety against adversarial attacks. However, we show that existing validation protocols fail to account for substantial distribution shifts inherent to red-teaming: diverse victim models exhibit distinct generation styles, attacks distort output patterns, and semantic ambiguity varies significantly across jailbreak scenarios. Through a comprehensive audit using 6642 human-verified labels, we reveal that the unpredictable interaction of these shifts often causes judge performance to degrade to near random chance. This stands in stark contrast to the high human agreement reported in prior work. Crucially, we find that many attacks inflate their success rates by exploiting judge insufficiencies rather than eliciting genuinely harmful content. To enable more reliable evaluation, we propose ReliableBench, a benchmark of behaviors that remain more consistently judgeable, and JudgeStressTest, a dataset designed to expose judge failures. Data available at: https://github.com/SchwinnL/LLMJudgeReliability.

Although large language models (LLMs) demonstrate expert-level medical knowledge, aligning their open-ended outputs with fine-grained clinician preferences remains challenging. Existing methods often rely on coarse objectives or unreliable automated judges that are weakly grounded in professional guidelines. We propose a two-stage framework to address this gap. First, we introduce HealthRubrics, a dataset of 7,034 physician-verified preference examples in which clinicians refine LLM-drafted rubrics to meet rigorous medical standards. Second, we distill these rubrics into HealthPrinciples: 119 broadly reusable, clinically grounded principles organized by clinical dimensions, enabling scalable supervision beyond manual annotation. We use HealthPrinciples for (1) offline alignment by synthesizing rubrics for unlabeled queries and (2) an inference-time tool for guided self-revision. A 30B parameter model that activates only 3B parameters at inference trained with our framework achieves 33.4% on HealthBench-Hard, outperforming much larger models including Deepseek-R1 and o3, establishing a resource-efficient baseline for clinical alignment.

We present a novel framework that improves the reliability of LLM judges by selectively augmenting LLM with auxiliary evaluation dimensions. Existing LLM judges often miss crucial evaluation dimensions because they fail to recognize the implicit standards underlying human assessments. To address this challenge, we propose the Auto-Prompt Ensemble (APE), an adaptive framework that automatically learns evaluation dimensions from its failure cases. APE incorporates a confidence-based ensemble mechanism to decide when to adopt the judgments from additional evaluation dimensions through a novel confidence estimation approach called Collective Confidence. Extensive experiments demonstrate that APE improves the reliability of LLM Judge across diverse standard benchmarks. For instance, APE enhances GPT-4o agreement rate on Reward Bench from 87.2% to 90.5% in the zero-shot setting. Overall, APE provides a principled approach for LLM Judge to leverage test-time computation, and bridge the evaluation gap between human and LLM judges.

Humanity's Last Exam (HLE) has become a widely used benchmark for evaluating frontier large language models on challenging, multi-domain questions. However, community-led analyses have raised concerns that HLE contains a non-trivial number of noisy items, which can bias evaluation results and distort cross-model comparisons. To address this challenge, we introduce HLE-Verified, a verified and revised version of HLE with a transparent verification protocol and fine-grained error taxonomy. Our construction follows a two-stage validation-and-repair workflow resulting in a certified benchmark. In Stage I, each item undergoes binary validation of the problem and final answer through domain-expert review and model-based cross-checks, yielding 641 verified items. In Stage II, flawed but fixable items are revised under strict constraints preserving the original evaluation intent, through dual independent expert repairs, model-assisted auditing, and final adjudication, resulting in 1,170 revised-and-certified items. The remaining 689 items are released as a documented uncertain set with explicit uncertainty sources and expertise tags for future refinement. We evaluate seven state-of-the-art language models on HLE and HLE-Verified, observing an average absolute accuracy gain of 7--10 percentage points on HLE-Verified. The improvement is particularly pronounced on items where the original problem statement and/or reference answer is erroneous, with gains of 30--40 percentage points. Our analyses further reveal a strong association between model confidence and the presence of errors in the problem statement or reference answer, supporting the effectiveness of our revisions. Overall, HLE-Verified improves HLE-style evaluations by reducing annotation noise and enabling more faithful measurement of model capabilities. Data is available at: https://github.com/SKYLENAGE-AI/HLE-Verified

Benchmarks measure whether a model is correct. They do not measure whether a model is reliable. This distinction is largely academic for single-shot inference, but becomes critical for agentic AI systems, where a single rephrased prompt can trigger cascading failures in multi-step execution. Yet this form of instability is not captured by existing evaluations. Hallucinations live in variance: they arise when semantically equivalent prompts activate inconsistent internal pathways, producing divergent outputs. Consistent but incorrect outputs reflect bias or missing knowledge; confident guessing reflects calibration failure. Neither constitutes hallucination under this definition. When error is variance-dominated, reducing redundant pathways improves reliability without adding knowledge. We formalize this through Semantic Stability (SS), measured via Paraphrase Consistency (PC@k): generate k paraphrases, greedy decode each, compute mode agreement. SS is a diagnostic for variance-driven unreliability, not a method for improving correctness. We show that a dense Qwen3-0.6B agrees with itself only 23.8% of the time; at 32% sparsity, agreement jumps to 55.9%. A phase diagram reveals the sweet spot where variance reduction outpaces bias accumulation, and regimes where stability collapses onto wrong answers.

As large language models (LLMs) continue to advance, reliable evaluation methods are essential particularly for open-ended, instruction-following tasks. LLM-as-a-Judge enables automatic evaluation using LLMs as evaluators, but its reliability remains uncertain. In this work, we analyze key factors affecting its trustworthiness, focusing on alignment with human judgments and evaluation consistency. Using BIGGENBench and EvalBiasBench, we study the effects of evaluation design, decoding strategies, and Chain-of-Tought (CoT) reasoning in evaluation. Our results show that evaluation criteria are critical for reliability, non-deterministic sampling improves alignment with human preferences over deterministic evaluation, and CoT reasoning offers minimal gains when clear evaluation criteria are present.

LLM-as-a-judge has become the dominant paradigm for grading model outputs at scale, yet the same model assigns systematically different scores when its output format changes (e.g., a 1-5 rating vs. a True/False label). Existing diagnoses of these format-induced inconsistencies stop at the input-output level. Using Position-aware Edge Attribution Patching (PEAP), we causally investigate the internal mechanism in Gemma-3, Qwen2.5, and Llama-3. We find that judgments across structured understanding and open-ended preference tasks share a sparse, generalized Latent Evaluator sub-graph in the mid-to-late multi-layer perceptrons (MLPs); zero-ablating it collapses judgment while preserving world knowledge in architecturally modular models. By structurally decoupling abstract judging from output formatting, we provide a mechanistic account of format-induced inconsistency on the open-weight models we study: a continuous judgment signal computed in the shared trunk is mapped through fragile, format-specific terminal branches, enabling format-independent preference to be isolated downstream of the requested output format. Our findings imply that benchmark-level reliability comparisons across formats are partially measuring formatter geometry rather than evaluation quality.

Large Language Models (LLMs) are increasingly employed as automated evaluators to assess the safety of generated content, yet their reliability in this role remains uncertain. This study evaluates a diverse set of 11 LLM judge models across critical safety domains, examining three key aspects: self-consistency in repeated judging tasks, alignment with human judgments, and susceptibility to input artifacts such as apologetic or verbose phrasing. Our findings reveal that biases in LLM judges can significantly distort the final verdict on which content source is safer, undermining the validity of comparative evaluations. Notably, apologetic language artifacts alone can skew evaluator preferences by up to 98\%. Contrary to expectations, larger models do not consistently exhibit greater robustness, while smaller models sometimes show higher resistance to specific artifacts. To mitigate LLM evaluator robustness issues, we investigate jury-based evaluations aggregating decisions from multiple models. Although this approach both improves robustness and enhances alignment to human judgements, artifact sensitivity persists even with the best jury configurations. These results highlight the urgent need for diversified, artifact-resistant methodologies to ensure reliable safety assessments.