LLM-as-a-Judge

Follow this topic Ask about this topic

Momentum

60 papers in the last four weeks, up 76% on the four weeks before. 0.6% of all new papers.

Jul 13Week of Sep 28

Latest papers 362

Jun 12, 2026cs.CV

Jury Duty: Calibration and Orientation Failures in MLLM-as-a-Judge Under Cultural Ambiguity

MLLM-as-a-Judge is conventionally validated by agreement with human annotations, but this metric is undefined when the human pool is culturally heterogeneous. We introduce VOIR DIRE, a multimodal benchmark of 626 culturally paired image--prompt artifacts spanning U.S. and mainland Chinese contexts across food, fashion, and architecture, with annotator pools that are within-pool reliable (a = 0.86/0.74) but cross-pool divergent on evaluation (Q1 r = -0.12). Across six MLLMs, the bias decomposes into two failures: a positivity-floor calibration failure (compressed scale use) and an orientation failure (default to one cultural norm). On this corpus, where contested items are sampled to split the two pools, the floor mechanically validates the more-permissive Chinese reading; persona prompting partially recovers calibration, but the orientation residual survives, evidence the tilt is not reducible to scale compression. Reference-pool in-context demonstrations deepen the orientation residual and inflate the high end rather than restoring use of the low end. Model origin adds a small additive tilt (~0.10 MAE) that is approximately invariant under demonstration. We recommend reporting alignment against each reference pool separately and treating cross-pool divergence as a judge property.

Daniel Lee, Harsh Sharma, Eunkyu Park +5

Salesforce AI Research · University of Colorado Boulder · Carnegie Mellon University +2

LLM-as-a-Judge Language Model Calibration Cultural Bias in Language Models Bias in LLM-as-a-Judge Multimodal Model Evaluation

Jun 12, 2026cs.CL

Does the Judge Prefer English? Evaluating Language-Switching Invariance in LLM-as-a-Judge

Large language models (LLMs) are now widely used as automatic judges for open-ended instruction-following evaluation. This practice is convenient, scalable, and often more semantically aware than reference-based metrics, but it also introduces a new reliability question: does a judge evaluate the quality of an answer, or does it also react to the language in which the comparison is presented? We propose Judge-LS, a lightweight meta-evaluation protocol that transforms LLMBar response-pair items into English, Chinese, and Chinese-English language-switched variants. A reliable judge should preserve its preference under label-preserving language transformations and should not prefer a language when two answers are translation-equivalent. We evaluate four API-accessible judges on the full 419-item LLMBar benchmark, producing 13,408 successful pairwise judgments. Across models, Chinese and language-switched presentations induce 10.7--14.4% preference flips relative to English, and all judges achieve their highest accuracy in English. However, translation-equivalent tie probes do not reveal a systematic English preference: most probes are judged as ties, and non-tie decisions more often favor Chinese. We add confidence intervals, paired significance tests, and an automatic transformation audit with a sensitivity analysis that excludes mechanically flagged high-risk variants. The experiment requires no model training, uses only API calls, and is feasible on modest local hardware.

Shaojie Yin

Shanghai Artificial Intelligence Laboratory Shanghai, China

Multilingual Language Model Evaluation LLM Evaluation LLM-as-a-Judge LLM Answer Verification Bias in LLM-as-a-Judge Language Model Robustness

Jun 11, 2026cs.LG

From Uncertain Judgments to Calibrated Rankings: Conformal Elo Estimation for LLM Evaluation

Evaluating new large language models typically requires costly human annotation campaigns at scale. LLM-as-a-judge offers a cheaper alternative, but judge scores carry systematic errors - such as position bias, self-preference, or intransitivity - that can strongly miscalibrate the resulting rankings. We quantify the resulting judge-human disagreement at two complementary levels. At the local level, we estimate per-battle uncertainty from the judge's own score differences by propagating calibrated win probabilities rather than hard labels into the Bradley-Terry procedure. This alone provides a drastic improvement to Elo estimation accuracy, bringing LLM-derived ratings within 17.9 Elo MAE of human-derived ones when averaged over 55 held-out models on LMArena. At the global level, we apply split conformal prediction to the residual gap between LLM-derived and human-derived Elo ratings across held-out models, producing prediction intervals with distribution-free marginal coverage guarantees that account for irreducible LLM-human disagreement. Together, these two layers yield a low-cost evaluation tool that provides developers with calibrated Elo estimates and honest uncertainty bounds, without access to large-scale human annotations. To facilitate reproducibility, we release our code at https://github.com/kargibora/SoftElo .

Bora Kargi, David Salinas

ELLIS Institute Tübingen · OpenEuroLLM

LLM Evaluation Pairwise Comparison LLM-as-a-Judge Pairwise Preference Evaluation Pairwise Ranking Language Model Calibration

Jun 10, 2026cs.DL

On the Limits of LLM-as-Judge for Scientific Novelty Assessment

LLMs are increasingly used to generate and judge scientific ideas. This makes novelty evaluation a central problem. Full idea evaluation is difficult because it often requires judging a method, its feasibility, and its empirical promise. We therefore study a cleaner upstream object: the research question (RQ). RQ generation is a prerequisite for scientific ideation, and RQs can be compared against questions pursued in real papers. We introduce RQ-Bench, a benchmark built from recent arXiv papers. For each paper, we reconstruct author-anchored RQs from its cited background, gaps, and contributions. These RQs are not the only valid questions for the same background. They are author-anchored reference points for testing novelty judgments. We evaluate model-generated RQs with standalone LLM judging, comparative LLM judging, and human expert evaluation. LLM judges consistently rate model-generated RQs as highly novel, producing a novelty mirage; in comparative evaluations, this preference becomes even stronger. Domain experts, however, reach the opposite conclusion and prefer the author-anchored reference questions. We further find that many generated RQs are narrow or source-bound, a dimension that LLM judges often miss unless explicitly tested. Overall, the contradictory novelty evaluations between LLM judges and human experts raise a serious concern about the reliability of using LLMs to assess the scientific novelty of research questions.

Soumitra Sinhahajari, Navonil Majumder, Soujanya Poria

DeCLaRe Lab, Nanyang Technological University

AI for Science LLM-as-a-Judge Research Idea Generation

Jun 9, 2026cs.CL

Catching One in Five: LLM-as-Judge Blind Spots in Production Multi-Turn Transaction Agents

LLM-as-judge is the default instrument for evaluating conversational agents, yet its reliability is almost always reported as agreement with human ratings, not recall of real defects. We study a deployed multi-turn food-and-beverage ordering agent and measure how many genuine quality problems its built-in LLM judge catches, using exhaustive human transcript review as ground truth. Across three batches the judge surfaces well under a quarter of human-confirmed systematic problems -- 2 of 9 patterns (22%) in one batch, and its operational gate flagged zero of 100 rounds in a batch where humans confirmed 23 distinct defects and 7 new cross-cutting patterns. Our blind-spot taxonomy shows the failure is structured, not random: the judge catches turn-local issues (a fabricated statistic, a wrong language) but misses cross-turn state issues (confirm-gate lockout, cart hallucination, escalation lockout, stale referents). The mechanism: the scoring rubric exposes only three coarse axes (intent, brand-voice, personalization) and has no category for the behavioural dimensions -- state-tracking, guardrails, recovery -- where most defects cluster. The failure is routing, not perception: 113 of 114 rounds whose raw judge note describes a confirm-gate or cart-state defect are scored "brand voice", and none reach an operational failure -- the gate is wired to hangs and hard assertions, not the rubric -- so the 0% is a routing-and-wiring failure, not blindness. The consequence for prevalence estimation is sharp: when the apparent defect rate is zero the Rogan-Gladen correction degenerates -- no signal can recover the true rate -- while where the gate reports a nonzero rate the same estimator implies a 3-6x undercount under our measured sensitivity. For production multi-turn agents, automated judging is a regression floor, not a substitute for human review.

Sawyer Zhang, Alexander Wang, Sophie Lei

Lumivate (Lumi)

LLM-as-a-Judge LLM Agent Evaluation Human-in-the-Loop Evaluation LLM Agent Reliability

Jun 8, 2026cs.AI

RealMath-Eval: Why SOTA Judges Struggle with Real Human Reasoning

While Large Language Models (LLMs) have achieved near-perfect performance in \emph{solving} high-school mathematics, their ability to \emph{evaluate} the diverse reasoning processes of real human students remains under-examined. To bridge this gap, we introduce \textbf{RealMath-Eval}, a rigorously annotated benchmark of 224 real-world exam responses from high schools. Our initial evaluation reveals that even state-of-the-art LLM judges struggle significantly on this task, exhibiting a high Mean Squared Error (

\sim

2.96) against expert human grading. To probe a plausible explanation, we contrast this performance with a control setting where the same judges evaluate synthetic LLM-generated solutions. We identify a stark

Evaluation Gap'': judges are considerably more accurate and consistent on synthetic text (MSE $\sim$1.17) but struggle to generalize to authentic student reasoning. Through semantic embedding analysis, we find that synthetic errors suffer from a

structural collapse'' into predictable, low-dimensional linear subspaces, whereas human errors form a more diverse error space. Furthermore, generative probability probes suggest that human reasoning involves significantly higher information-theoretic surprisal, indicating that student reasoning transitions are more out-of-distribution for current models. Finally, we find that surface-level style transfer fails to close this gap. Our findings suggest that current LLM evaluation pipelines relying heavily on synthetic data may not adequately capture the diversity of authentic student mathematical reasoning.

Yiteng Mao, Kenan Xu, Yijia Lyu +3

University of Wisconsin–Madison · East China Normal University · New York University +2

Mathematical Reasoning Benchmarks LLM Evaluation LLM-as-a-Judge Automated Grading Synthetic Data Evaluation

Jun 8, 2026cs.CL

Reasoning without Gold Standards: A Proxy-Judge Theory of Autoformalization

Complex reasoning tasks increasingly require systems to produce outputs whose correctness cannot be judged by exact match against a single reference. Autoformalization (AF) is a representative example; it asks a model to translate informal mathematical or logical reasoning into a formally checkable object, yet expert-validated formalizations do not scale beyond toy cases and a single informal argument can admit many valid formal renderings. Progress therefore depends on whether partial, structured proxies can substitute for exact references. We introduce a reference-free proxy-judge framework for AF that replaces gold-standard matching with a vector of per-axis property checks. The framework organizes the proxy along three structural scopes that cover global properties of the elicited object, per-module properties internal to its sub-components, and cross-domain properties that re-align it to the informal source, and aggregates each axis into a verdict vector. The vector drives a reflective refinement loop in which a violated coordinate routes the controller to a matching repair target, so each iteration changes only what is judged wrong. Under bounded judge noise, the expected intrinsic gap contracts geometrically to a noise-dependent plateau. Across seven formalization backbones on miniF2F, ProofNet, e-SNLI, and ProntoQA, refinement consistently lifts Pass Rate over the single-shot ICL baseline, and the per-axis proxy outperforms a matched scalar proxy on benchmarks where the baseline has room to improve. Structured proxy judgments therefore provide both a practical refinement signal and a theoretical handle on convergence when exact references are unavailable.

Lei Xu, Xin Quan, André Freitas

1Idiap Research Institute, Switzerland · 2École Polytechnique Fédérale de Lausanne (EPFL), Switzerland · Department of Computer Science, University of Manchester, United Kingdom +1

LLM Self-Refinement LLM-as-a-Judge Reference-Free Evaluation Automated Evaluation Autoformalization

Jun 8, 2026cs.AI

Correct Looks Better: Pairwise Comparisons Reveal Accuracy Rankings

Pairwise comparisons combined with aggregation methods like Elo have become central to evaluating generative models, yet concerns remain that they reward superficial stylistic cues or display judge biases. In a more positive turn, we show that model rankings from pairwise comparisons strongly agree with ground-truth-based accuracy rankings when such ground truth is available for comparison. By converting five well-known benchmarks into free-form generative evaluations, we find that Elo rankings achieve a Spearman correlation above 0.9 with accuracy rankings and substantially outperform direct evaluation when the judge is weak. Furthermore, style and judge bias have only minor effects on model rankings, despite most judgments occurring on pairs where both candidate answers are correct (or incorrect). On such pairs, we find that repetition after the final answer (echo) is a causal driver of judge preference.

Mina Remeli, Moritz Hardt

Max Planck Institute for Intelligent Systems, Tübingen, Germany, Tübingen AI Center

Pairwise Comparison LLM-as-a-Judge Language Model Generation Evaluation Pairwise Preference Evaluation Pairwise Ranking Automated Evaluation+1

Jun 8, 2026cs.AI

Reliable to Expressive: A Curriculum for Rubric-Following Safety Judges

Safety judges are increasingly deployed to evaluate model outputs against evolving criteria, yet recent meta-evaluation work shows they remain brittle under prompt and rubric variation, with false negative-rate swings of up to 0.24 reported for stylistic perturbations alone. We argue that safety judgment is fundamentally a rubric-following problem: a robust judge must apply the given evaluation criteria consistently across rubric formulations rather than memorize one specific template. We propose a training strategy that combines (i) instance-conditioned dynamic rubrics generated from prompt-response-label triples to expose the judge to the variability of evaluation criteria, and (ii) a reliable-to-expressive curriculum that begins with clean fixed-rubric supervision and progressively introduces noisier dynamic-rubric data. We evaluate on a single human-labeled set under three contrasting rubric prompts (HarmBench-style, ShieldGemma-style, and a domain-specific rubric). Our 12B curriculum judge achieves 94.12-94.88% accuracy across the three rubrics with a cross-rubric range of only 0.76, outperforming general-purpose LLMs, dedicated safety classifiers, and reasoning-oriented judges up to 30B in both peak accuracy and stability. An ablation shows that naively mixing dynamic rubrics into SFT increases cross rubric variance (1.44 -> 3.60); only the curriculum schedule recovers and improves on the fixed rubric baseline (variance 0.76).

Yongtaek Lim, Hyeji Choi, Minwoo Kim

AI Safety Team, DATUMO.INC, Seoul, South Korea.

LLM-as-a-Judge LLM Safety Evaluation Curriculum Learning Rubric-Based Evaluation Language Model Robustness

Jun 8, 2026cs.AI

ComplexConstraints and Beyond: Expert Rubrics for RLVR

Evaluation protocols can lag behind LLM capabilities. Programmatically verified benchmarks cover narrow surface constraints, whereas real-world instruction following and agentic workflows require judging semantic, contextual, and policy-dependent behavior. We study expert-curated rubric-based evaluation as a unified mechanism for measurement and reinforcement-learning rewards across two settings: complex instruction following and enterprise agentic tasks. We identify rubric-design choices that affect reward quality, including maximum viable atomicity, intent-aware criterion design, and LLM-judge calibration. We introduce ComplexConstraints, an expert-curated instruction-following suite comprising a public 75-prompt benchmark with 1,559 rubric criteria and a disjoint 1,000-prompt training set, with 10-40 atomic criteria per prompt. Empirically, rubric rewards improve training in both fixed task datasets, such as ComplexConstraints, and stateful RL environments, such as CoreCraft. Training a 4B model on ComplexConstraints improves mean criterion pass rate by +15.5 pp on a held-out split, bringing it within 0.5 pp of the untrained baseline of a roughly 60x larger Qwen3 model, and the gains transfer to external benchmarks the model never saw during training: +8.4 pp on AdvancedIF and +10.1 pp on MultiChallenge. In CoreCraft, rubric-reward RL likewise transfers to out-of-distribution benchmarks (+4.5 pp BFCL, +7.4 pp tau^2-Bench, +6.8 pp Toolathlon). These results show that expert-authored rubrics provide effective evaluation targets and scalable reward signals for improving LLM instruction following and agentic behavior.

Sushant Mehta, Liudas Panavas, Suhaas Garre +1

Surge AI

LLM-as-a-Judge Rubric-Based RL Instruction Following Reinforcement Learning with Verifiable Rewards LLM Agent Training Rubric-Based Evaluation+1

Jun 7, 2026cs.AI

AgentTrust: A Self-Improving Trust Layer for AI-Agent Actions

AI agents increasingly take consequential actions -- shell commands, cloud operations, and arbitrary tool-calls -- so a trust layer must decide, per action, whether to allow, warn, block, or escalate. We argue that the right way to reason about such a layer is by threat type. Lexical (fixed-signature) threats, where danger lives in a stable token, are decidable by deterministic rules; semantic (intent-dependent) threats, where a benign and a malicious action share the same surface, are out of reach for rules by construction. We make this concrete with a negative proof: a determined, hand-authored cloud rule pack lifts held-out accuracy only 48 to 56% overall and moves the semantic categories by 0pp (data_db 29 to 29, observability 59 to 59, supply_chain 50 to 50), while a strong LLM judge carries exactly those categories. We give the judge a self-learning capability: on a corpus that is mainly semantic attacks it nearly doubles rule accuracy (48% to 83.6-85.2%) with near-zero false-blocks, and this holds across two model providers. We turn this into a self-improving dual-store system: the judge distills a growing deterministic rule floor on lexical threats (cheaper over time) and feeds a guarded RAG memory on semantic threats (a verdict-cache fails -- surface-twins collapse to ~58% -- so a corroboration guard lifts semantic accuracy +13pp, 70 to 84). The result is what sets AgentTrust v2 apart from its static v1 predecessor: a trust layer that self-evolves from its own stream of decisions -- cheaper on the lexical class (it distils its own rules) and smarter on the semantic class (it accrues guarded precedent), while never hard-blocking a benign action. An end-to-end online replay shows the judge-call rate falling (50% to 44%) and judge-domain accuracy rising (71% to 80%), with 0 benign hard-blocks across 45,000 actions.

Chenglin Yang

Independent Researcher

AI Agent Reliability LLM-as-a-Judge AI Agent Security LLM Agent Security Runtime Enforcement for AI Agents Tool-Use Authorization+1

Jun 7, 2026cs.SE

Impacts of Histories and Models on LLM Grading: A Study in Advanced Software Engineering Courses

Graduate-level research reading report assessment creates a substantial labor burden for educators. While large language models (LLMs) hold great potential for automating academic grading, their reliability for this specialized task remains understudied, particularly regarding grading consistency, the lack of which represents a primary obstacle to educational fairness. This paper proposes a human-aligned LLM-assisted grading workflow and presents a case study based on 180 student submissions from a graduate advanced software engineering course. We evaluate two mainstream LLMs, Grok and GPT, in terms of grading consistency and alignment with human scores. We find LLMs exhibit distinct levels of intra-model consistency and significant inter-model grading inconsistencies, while simple ensemble approaches cannot improve alignment with human evaluation. Critically, continuous interaction history drives systematic drift in models' grading standards away from human expert scores. Our findings demonstrate LLMs' potential in reducing grading workload for educators in graduate education, while highlighting that indiscriminate LLM grading may introduce systemic unfairness, suggesting that specific operational practices are required to mitigate such disparities.

Qilin Zhou, Zhuo Wang, Yue Li +1

City University of Hong Kong, Hong Kong SAR, China

LLM Evaluation LLM-as-a-Judge LLM Prompting AI in Education Automated Grading

Jun 6, 2026cs.CY

Contemporary AI lacks the imagination to diverge or negate in science

Bold projections that artificial intelligence will accelerate scientific discovery have raced ahead of evidence from working scientists, and the field still lacks large-scale, scientist-in-the-loop tests of these claims. Here we mount the largest such evaluation to date and map what AI cannot yet do for science. We invited authors of 121,640 recent preprints across biology, medicine, chemistry, and the social sciences to judge ideas that large language models (LLMs) generated from the context and puzzles of their own papers. 6,749 scientists returned 25,139 sets of ratings on novelty, empirical feasibility, probability of being true, and favorability of adoption. Three patterns emerge. First, non-reasoning LLMs collapse into a narrow "hivemind" of similar ideas; reasoning models roam a wider hypothesis space, yet no model class spontaneously proposes null hypotheses -- a move humans make more freely. Second, scientists reward ideas that resemble their own and prize probability over novelty, though social scientists tolerate risk more readily than life scientists. Senior social scientists are the harshest critics, and their skepticism is well-earned: LLMs falter most in pluralistic fields like the social sciences that demand context-aware interpretation and evolving theories. Third, automated evaluators on which the community currently relies -- LLM-as-a-judge, artificial metrics, and even state-of-the-art (SOTA) models -- agree only weakly with expert judgment, and retrieval augmentation and scientist persona prompting yield only marginal gains. A Qwen3-14B reward model we post-trained on human ratings captures field taste nuances, beats SOTA models by up to 27%, and closes the gap to the inter-rater consistency of independent peer reviewers. For all the hype, today's scientific AI still represents a collaborator whose imagination, outputs and judgment benefit from human grounding.

Honglin Bao, Siyang Wu, Xiao Liu +3

Data Science Institute, University of Chicago · Knowledge Lab, University of Chicago

LLM-as-a-Judge Scientific Reasoning Scientific Hypothesis Generation Human-in-the-Loop Evaluation

Jun 6, 2026cs.AI

Online Agent-as-a-Judge: Situation-Generating Evaluation for Interactive Agents

Evaluating LLM-powered interactive social agents is challenging because socially relevant behaviors depend not only on isolated outputs, but also on prior interactions, social roles, and downstream actions. Existing methods typically allow a target agent to act freely in an environment and then score the resulting trajectory. However, this passive setup can miss capabilities that only become observable under specific social circumstances; for example, conflict handling may remain untested if no disagreement arises. We propose Online Agent-as-a-Judge, a situation-generating evaluation framework for interactive social agents. Online Agent-as-a-Judge deploys an in-world evaluator agent that interacts with the target agent through the environment's native dialogue and action protocol, actively eliciting situations relevant to the evaluation criteria. The resulting trajectories provide evidence for assessing both immediate responses and subsequent behavior. In a life-simulation environment with

32

designer-authored social criteria, Online Agent-as-a-Judge improves criteria coverage and agreement with human labels, yielding more reliable evidence-grounded evaluations of behaviors that passive methods can leave unobserved.

Hyogon Ryu, Jeonghwan Kim, Yewon Lim +3

KRAFTON

LLM-as-a-Judge LLM Agent Evaluation Agent Evaluation

Jun 6, 2026cs.CL

When Languages Disagree: Self-Evolving Multilingual LLM Judges

Multilingual LLM-as-a-judge is widely used to evaluate model outputs across languages, but suffers from cross-lingual inconsistency (Fu and Liu, 2025). Existing methods typically treat this inconsistency as noise and mitigate it through voting or aggregation. In this work, we instead show that multilingual inconsistency can provide complementary evaluation signals. Our oracle analysis finds that sampling judgments across languages yields a higher performance upper bound than single-language judging, indicating that different languages potentially include complementary judgments. Motivated by this finding, we propose SEMJ, a self-evolving multilingual judge that leverages cross-lingual inconsistency for iterative refinement. SEMJ constructs multilingual variants of each input, collects independent judgments and rationales, and feeds inconsistent outputs back for self-reflection and re-evaluation. Experiments on multiple benchmarks show that SEMJ consistently outperforms voting and reflection baselines in both accuracy and cross-lingual consistency. Further analysis shows that inconsistency triggers useful re-evaluation, which improves judgment quality.

Xiyan Fu, Wei Lu

Nanyang Technological University

Multilingual Language Model Evaluation LLM-as-a-Judge Language Model Self-Improvement Cross-Lingual Consistency in Language Models

Jun 5, 2026cs.AI

Safety is Contextual, LLM-Judges Are Not: Navigating the Rigid Priors of Evaluators

LLMs-as-judges are the primary way to evaluate safety at scale. Despite their importance, LLM-judges themselves are rarely evaluated beyond human agreement in simple, static benchmarks. We therefore investigate two under-explored but crucial properties of LLMs-as-judges: their sensitivity to in-context information, and their steerability to differing safety definitions, which may not align with their internal safety priors. We evaluate the safety judging abilities of 13 generalist LLMs and safety-specific judges, and investigate the impact of novel in-context information and changing safety definitions. We find that while LLM-judges can learn from new information, they are broadly unlikely to update their evaluations if the context or safety definition departs from their prior.

Anissa Alloula, Federico Licini, Ava Batchkala +1

University of Oxford · Cohere

LLM-as-a-Judge In-Context Learning LLM Safety Evaluation

Jun 5, 2026cs.SE

Cherry-pick Override: Unsafe Directional Commitment in LLM Judges under Mixed Evidence

LLM judges increasingly turn verdicts into system commitments. Under mixed evidence (claims with both supporting and refuting sources) this is unsafe: when the schema exposes CONFLICTING as the authorized non-directional verdict, returning SUPPORTS/REFUTES is an unauthorized directional commitment, a failure we name Cherry-pick Override (CCO). We define CCO under an explicit task contract and report it with a same-denominator diagnostic protocol paired with matched-coverage bootstrap and an apples-to-apples random-veto null. On AVeriTeC's Conflicting subset (N_C = 150), three-option judges return a directional verdict on more than 84% of mixed-evidence claims; under the typed schema, three-judge majority voting amplifies direction-on-conflict on AVeriTeC (0.887 vs. 0.840; 95% CI [+0.013, +0.080]) but does not replicate on VitaminC-Mixed. Walking an intervention ladder of common single-channel fixes (typed vocabulary, panel aggregation, confidence thresholding, validator-only filtering), each leaves a distinct residual failure: panel aggregation suppresses single-judge CONFLICTING dissent in 48% of CCO cases; the panel is well-calibrated for direction (ECE = 0.07 on pure-S/R) so confidence cannot operationally separate CCO from correct directional commits; validator-as-classifier nearly halves pure-evidence accuracy. A minimal two-channel reference probe reaches operating points neither single channel reaches; under the random-veto null its promotion to CONFLICTING is structurally targeted on AVeriTeC (empirical p < 1/2001) and weaker but in the same direction on VitaminC-Mixed, a selectivity result rather than a magnitude one. We argue for an external commitment-control layer that separates verdict generation from commitment authorization, using structural evidence and confidence as orthogonal channels and NO-COMMIT as a routed controller state.

Haoran Xu

University of Glasgow

LLM-as-a-Judge Evidence-Grounded Reasoning LLM Abstention Language Model Auditing

Jun 5, 2026cs.CL

SLMJury: Can Small Language Models Judge as Well as Large Ones?

Large language models (LLMs) are widely used as judges for evaluating model outputs, but their high cost, latency, and opacity limit scalability. We introduce SLMJury, a framework for evaluating small language models (SLMs) as judges across two paradigms: closed-ended binary correctness and open-ended quality scoring. We benchmark 16 SLM judges (0.6B-14B parameters) from four model families across ten benchmarks: eight closed-ended tasks spanning mathematical, scientific, and general reasoning (N=64,824 judgments per configuration), plus SummEval and MT-Bench for summarization and conversational scoring. We formalize judging as a budget-conditioned function and study five dimensions. Four findings emerge. (1) The overthinking effect is domain-dependent: for most judges quick 10-token verdicts match or beat extended reasoning on mathematical judging (by 2-7% where they help), while reasoning wins on general tasks by up to 23%. (2) Domain generalization separates model families, with math-to-general accuracy gaps ranging from under 10% to nearly 40%. (3) Closed-ended and open-ended judging draw on different capabilities: the best binary judge (Phi-4) drops to rank 9 on MT-Bench, while reasoning-trained models invert this ordering. (4) Under the Reflect-Critique-Refine (RCR) debate protocol, multi-agent debate degrades accuracy across all tested configurations, whereas the top judges resist six adversarial personas with <=0.55% variance. Reliable automated evaluation does not require large proprietary models, yet no single SLM dominates. The leaderboard is available at https://anishh15.github.io/SLMJury/, and our framework code and pip package are publicly available at https://github.com/anishh15/SLMJury and https://pypi.org/project/slmjury/.

Anish Laddha, Nitesh Pradhan, Gaurav Srivastava

Department of Computer Science and Engineering, LNMIIT, Jaipur, India · Department of Computer Science, Virginia Tech, Blacksburg, VA, USA

LLM-as-a-Judge Language Model Generation Evaluation Small Language Models Automated Evaluation Language Model Robustness

Jun 5, 2026cs.CL

Beyond Rubrics: Exploration-Guided Evaluation Skills for Reward Modeling

Open-ended reward modeling requires judges that can follow subtle, domain-specific preferences when verifiable answers are unavailable. Existing rubric-based methods often address this by generating criteria online for each query, but the extra generation step can add inference overhead and produce rigid or misaligned guidance. We introduce Eval-Skill, an exploration-guided method that synthesizes reusable evaluation skills for reward modeling and reframes reward guidance as context evolution rather than parameter training or per-query rubric generation. Using only 100 cases per domain for skill evolution, Eval-Skill synthesizes reusable domain-level evaluation skills through two progressive stages, workflow generation followed by principle generation, with exploration and selection interleaved across both stages. Once generated, a skill is directly injected into the judge context. Across multiple RM benchmarks, Eval-Skill consistently improves diverse judge backbones; on RewardBench 2, it yields significant gains over vanilla judging for each main backbone (+13.44% for Qwen3-8B, and 18.51% for DeepSeek-V4-Flash). Further analyses of evolution-time scaling, generalizability, and transferability show that compact evaluation skills offer an efficient new paradigm for LLM-based evaluation. Code is available at https://github.com/xing-stellus-yue/Eval-Skill.

Xing Yue, Linjuan Wu, Daoxin Zhang +2

Zhejiang University · Xiaohongshu Inc.

Reward Modeling LLM Evaluation LLM-as-a-Judge Reward Model Evaluation

Jun 4, 2026cs.MA

MADRAG: Multi-Agent Debate with Retrieval-Augmented Generation for Training-Free Analytic Essay Scoring

We present MADRAG, a training-free framework for analytic essay scoring that combines multi-agent reasoning with retrieval-augmented grounding. Unlike standard LLM-as-judge approaches, which are prone to bias and unstable scoring, MADRAG decomposes evaluation into an interactive process: an Advocate identifies strengths, a Skeptic critiques weaknesses, and a Judge aggregates their arguments into a final score. Crucially, the Judge is augmented with rubric-aligned exemplar retrieval, enabling calibration through comparison with scored examples. Our results show that MADRAG significantly outperforms prompt-based baselines while approaching the performance of supervised systems without requiring task-specific training. Ablation studies demonstrate that retrieval drives calibration gains, while debate improves reasoning on higher-level traits. Our findings highlight the complementary roles of structured interaction and external memory in reliable LLM-based evaluation.

Ali Keramati, Shiyuan Zhou, Sharad Mehrotra +1

University of California, Irvine

Multi-Agent LLM Systems LLM-as-a-Judge Multi-Agent Debate Automated Essay Scoring Rubric-Based Evaluation

Jun 4, 2026cs.CL

EDIT: Evidence-Diagnosed Intervention Training for Rule-Faithful LLM Grading

Reliable rubric grading requires more than accurate score prediction. Each judgement must be grounded in the mark scheme and evidence from the student answer. Existing credit-assignment and intervention methods, primarily designed for self-contained reasoning tasks such as mathematics reasoning, struggle in this setting because they do not identify where grading reasoning goes wrong or how the model's belief about the final mark changes during reasoning. We propose Evidence-Diagnosed Intervention Training (EDIT), a two-phase framework for training more rubric-faithful LLM graders. First, EDIT-SFT locates problematic reasoning steps using internal model signals: posterior belief over the final mark and input-grounding scores. It then revises only these local steps with help from a rubric checklist. Second, EDIT-RL calibrates the grader with belief-guided reward shaping, penalising large harmful belief drifts while still allowing helpful exploration. Experiments on two real-world, multi-subject grading benchmarks demonstrate that EDIT consistently outperforms strong supervised fine-tuning and reinforcement learning baselines on both in-domain and out-of-domain splits, with ablation studies confirming that internal-state diagnostics drive these gains. Under deterministic rubric-edit interventions, EDIT-SFT is the most rule-responsive of all evaluated systems, and EDIT-RL largely retains this responsiveness while improving accuracy.

Zhihao Wu, Linhai Zhang, Taiyi Wang +4

1King’s College London · University of Cambridge · 3AQA +1

LLM-as-a-Judge Rubric-Based Evaluation LLM Post-Training Automated Grading

Jun 4, 2026cs.LG

Elmes*: Automated Construction of Fine-Grained Evaluation Rubrics for Large Language Models in Long-Tail Educational Scenarios

Evaluating large language models (LLMs) for education requires measuring how models teach, not only what they know. Existing benchmarks emphasize domain-general correctness or depend on manually designed rubrics that scale poorly to long-tail pedagogical scenarios. We introduce Elmes*, an end-to-end framework for constructing, refining, and applying fine-grained scenario-specific rubrics. Elmes* combines a declarative multi-agent engine for teacher--student--judge interactions with SceneGen, a self-evolving module that co-optimizes evaluation criteria and test data from expert-defined pedagogical dimensions. Using Elmes*, we build Edu-330, covering 330 scenarios across 11 subjects, 3 grade bands, and 10 task types, with over 1{,}000 second-level indicators. Experiments on Edu-330 and four expert-authored gold-standard scenarios show that educational capability is multidimensional: top-tier LLMs differ mainly in creativity and values integration, knowledge-strong models may fail at Socratic scaffolding, and the education-specialized InnoSpark achieves the best human-evaluated average score. LLM judges preserve human-comparable rankings with much lower scoring variance, but exhibit judge-specific biases such as self-preference. Ablations show that expert-scored few-shot anchoring improves human--LLM alignment, while reasoning enforcement and greedy decoding are model-dependent. Elmes* thus provides scalable diagnostic infrastructure for pedagogically grounded LLM evaluation.

Tao Liu, Ye Lu, Ruohua Zhang +4

1Shanghai Institute of AI for Education, East China Normal University, Shanghai 200062, China · School of Computer Science and Technology, East China Normal University, Shanghai 200062, China · 3Shanghai Innovation Institute, Shanghai 200231, China

LLM Evaluation LLM-as-a-Judge Language Model Generation Evaluation Rubric-Based Evaluation

Jun 3, 2026cs.AI

Stability vs. Manipulability: Evaluating Robustness Under Post-Decision Interaction in LLM Judges

LLM-as-judge evaluation is widely used in benchmarking pipelines, where model outputs are compared and ranked using automated evaluators. These pipelines typically assume that judgments are stable properties of fixed inputs. We show that this assumption does not hold under interaction. We study post-decision manipulability: the extent to which an evaluation outcome can be altered through subsequent conversation with the judge after an initial decision has been made. Across controlled experiments on MT-Bench and AlpacaEval, we find that LLM judges are highly stable under repeated and neutral reevaluation, yet become substantially reversible under targeted post-decision challenge. An anti-baseline challenge protocol shows that stable judgments can be overturned through motivated interaction, while a counterbalanced target-validation protocol separates this reversibility from net target-directed steering. These reversals have practical consequences: they can degrade agreement with human preferences, shift benchmark rankings, and produce harmful evaluation changes despite high self-reported confidence. Authority framing is especially destabilizing, and revised judgments are often accompanied by low-overlap justifications, suggesting post hoc rationalization rather than reliable error correction. We introduce the Evaluation Robustness Score (ERS) to quantify interactional robustness by combining reversal susceptibility with counterbalanced directional effects. Our findings identify post-decision interaction as a distinct failure mode for LLM-as-judge evaluation and motivate evaluation protocols that measure not only static agreement, but robustness under challenge.

Srimonti Dutta, Akshata Kishore Moharir

WAI USA Research Labs

LLM-as-a-Judge Language Model Robustness

Jun 3, 2026cs.LG

Statistically Reliable LLM-Based Ranking Evaluation via Prediction-Powered Inference

With PRECISE, we extended Prediction-Powered Inference to produce bias-corrected estimates of ranking evaluation metrics by combining a small human-labeled set with a large LLM-judged set. PPI is provably unbiased regardless of the LLM judge's error profile. We make it applicable to hierarchical metrics like Precision@K, where annotations are per-document but the metric is per-query, by reducing the output-space computation from O(2^|C|) to O(2^K). On the ESCI benchmark, augmenting 30 human annotations with Claude 3 Sonnet judgments reduces the standard error of Precision@4 estimates from 4.45 to 3.50 (a 21% relative reduction). In a production system, our framework correctly identified the best of three system variants from 100 human labels and 2 hours of domain-expert annotation; A/B testing confirmed this ranking with +407 bps in daily sales.

Abhishek Divekar

Amazon

LLM-as-a-Judge IR Evaluation Prediction-Powered Inference

Jun 3, 2026cs.CL

Self-Evaluation Is Already There: Eliciting Latent Judge Calibration in Base LLMs with Minimal Data

Large language models are increasingly evaluated by other models, raising a natural question: can a model predict how a judge will score its own output? We find that the ability is largely present before any targeted training: prompted few-shot, a base model already predicts an external judge's multi-attribute quality scores on open-ended responses well above chance across three benchmarks. We introduce Self-Evaluation Elicitation (SEE), a method that surfaces this latent ability through a short cycle comprising a calibration-coupled reinforcement learning phase that improves the answer and predicts the judge, followed by a masked distillation phase that sharpens the prediction while leaving the answer untouched. From 160 unique examples, roughly 31x fewer than a reinforcement learning baseline, SEE improves held-out calibration across three benchmarks while preserving answer quality. The elicited self-evaluation is sharply localized within the model's own token distribution and stable across judges it was never trained against, indicating a transferable notion of quality rather than a single judge's preference. These results reframe judge-aligned self-evaluation as a problem of elicitation rather than acquisition.

XiuYu Zhang, Yi Shan, Junfeng Fang +1

National University of Singapore · Beijing University of Technology

LLM Evaluation LLM-as-a-Judge Language Model Generation Evaluation Language Model Self-Assessment Language Model Calibration Evaluation Awareness in Language Models

Jun 3, 2026cs.LG

Reproducing, Analyzing, and Detecting Reward Hacking in Rubric-Based Reinforcement Learning

Rubric-based reinforcement learning (RL) uses an LLM-as-a-Judge (LaaJ) to score model outputs according to rubrics as rewards. However, policy models may exploit latent biases in the judge, leading to reward hacking and ineffective or unsafe training outcomes. In real-world rubric-based RL, such hacking behaviors are often subtle and entangled with multiple judge biases, making them difficult to analyze, detect, and mitigate. In this paper, we introduce CHERRL, a Controllable Hacking Environment for Rubric-based RL. By injecting known biases into LaaJ, CHERRL enables stable reproduction of reward hacking, explicit observation of reward divergence, and identification of hacking onset. This provides a clean experimental testbed for studying the mechanisms and mitigations of reward hacking in rubric-based RL. To demonstrate its utility, we analyze different judge biases from the perspectives of discoverability and exploitability, and explore an agent for automatically detecting reward hacking onset from training logs. The code and environment are publicly available at https://github.com/THUAIS-Lab/CHERRL.

Xuekang Wang, Zhuoyuan Hao, Shuo Hou +3

Tsinghua University · Harbin Institute of Technology, Shenzhen · Xi’an Jiaotong University

Reward Hacking RL for Language Models LLM-as-a-Judge Reinforcement Learning Rubric-Based RL

Jun 3, 2026cs.AI

AICompanionBench: Benchmarking LLMs-as-Judges for AI Companion Safety

As AI companion platforms such as Replika and Character.AI rapidly grow, concerns about unsafe human-AI interactions have intensified. This study introduces AICompanionBench, to our knowledge the first publicly available benchmark dataset of human-AI companion conversations annotated with fine-grained safety risk categories. The dataset contains 2,123 real-world Replika conversations collected from Reddit and annotated through human-AI collaboration across nine categories: sexual behavior, antisocial behavior, physical aggression, verbal aggression, substance abuse, self-harm and suicide, control, manipulation, and no-harm. Using this benchmark, we evaluate 20 state-of-the-art open-source and closed-source LLMs under an LLM-as-judge framework for detecting unsafe interactions. Results show substantial variation in model performance, with stronger models achieving high overall accuracy but still struggling with nuanced categories such as manipulation, as well as benign conversations that are incorrectly identified as harmful. Our findings suggest that while current LLMs can effectively detect explicit harmful content, they remain limited in identifying implicit unsafe interactions. Overall, our work contributes a new benchmark dataset for AI companionship safety research and offers insights into monitoring AI companion systems using LLMs. The dataset is publicly available at: https://github.com/anonymousresearcher2026/AICompanionBench/blob/main/AICompanionBench.xlsx

Yanjing Ren, Reza Ebrahimi, TengTeng Ma

The School of Information Systems University of South Florida Tampa, US

LLM Safety Benchmarks LLM-as-a-Judge AI Companions AI Safety Evaluation LLM Safety Evaluation AI Agent Safety Benchmarks+1

Jun 2, 2026cs.AI

The Saturation Trap and the Subjectivity of Intervention Timing: Why Affect-Based Triggers and LLM Judges Fail to Time Interventions on Autonomous Agents

As autonomous AI agents move from conversational systems to long-horizon software execution, runtime safety layers that decide when to interrupt an agent have become essential. We study this timing problem using a continuous 18-dimensional affective-dynamics engine (HEART) as a diagnostic probe, evaluating four intervention trigger families - absolute state thresholds, composite state-action patterns, regex reasoning-feature extraction, and zero-shot LLM-as-judge - against human-annotated intervention points on SWE-bench-Verified debugging traces. We report three findings. First, a State Saturation Trap: agents show no recovery signal under sustained difficulty, so modeled frustration quickly crosses the threshold and stays at its maximum, converting threshold-on-state triggers from moment detectors into near-constant indicators that fire on 39-83% of actions across five trajectories. Second, a capability-and-context floor for LLM judges: a small model (gpt-5.4-mini) never fires, while frontier and cross-vendor models escape the zero-firing floor only with full-trajectory context, and even then reach only F1 0.17-0.40 at up to 90x the cost. Third, and most importantly, the supervised target is not reproducible among humans: three trained annotators using one rubric on a 56-action trajectory agree on where to intervene only slightly above chance (location Krippendorff's alpha = +0.047; best pairwise Cohen's kappa = +0.349) and not at all on intervention type (pause degenerate; clarify below chance; reflect only alpha = +0.226). We conclude that intervention timing is a low-reliability construct, making single-annotator F1 an unsuitable optimization target. Our contribution is the joint mapping of this problem across human inter-rater reliability, four detector architectures, a cross-model LLM-judge sweep, and a reproduced saturation effect, rather than any single detector's accuracy.

Manvendra Modgil

LLM-as-a-Judge Human-in-the-Loop Evaluation AI Agent Monitoring Runtime Enforcement for AI Agents Inference-Time Intervention

Jun 2, 2026cs.CL

AI Rater Discrimination Depends on Scoring Protocol in Complex Clinical Decision-Making

Clinical AI evaluation increasingly delegates scoring to large language models (LLMs) acting as AI raters, yet their scoring behavior across evaluation conditions has not been quantitatively characterized. We address this gap through a factorial study of AI rater behavior in adult type 2 diabetes (T2D) pharmacotherapy at 12-month outpatient follow-up, a clinical task involving complex decision-making operationalized across seven evaluation questions. Four open-source LLMs served simultaneously as clinical decision support system (CDSS) models and AI raters. Each CDSS output was scored under two scoring protocols: a rubric-anchored Gold Rubric (GR) protocol incorporating a patient-specific rubric, and a rubric-free Non Gold Rubric (Non-GR) protocol. Linear mixed effects models crossed the scoring protocol factor with five design factors -- CDSS model, CDSS prompt configuration (document-referenced generation [DRG] vs.\ Baseline), rater model, prompt character, and prompt type -- and estimated main effects together with their protocol interactions. Across all questions, AI raters yielded consistently higher scores within a very narrow range (74--78 points on average) under Non-GR compared to those under GR (7.69 to 49.64 points lower mean scores; 1.68 to 3.67 times wider interquartile ranges). Within each question, GR amplified the AI rater's discrimination between DRG and Baseline CDSS outputs by factors of 1.76 to 5.10, while also revealing substantial behavioral variation across rater models that Non-GR suppressed. These findings support rubric anchoring as the scoring protocol that preserves discriminative power in clinical AI evaluation; rubric-free scoring cannot substitute when questions require patient-specific or jurisdiction-specific criteria that rater models cannot infer from parametric knowledge alone.

Sangwon Baek, Kyu Yeon Hur, Kyunga Kim

Asclep Korea Inc., Changwon, Republic of Korea · Center for Data Science, New York University, New York, USA · Division of Endocrinology and Metabolism, Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea +1

LLM-as-a-Judge Clinical Decision Support Clinical Decision-Making Rubric-Based Evaluation Clinical Language Model Evaluation

Jun 2, 2026cs.CL

SenseJudge: Human-Centric Preference-Driven Judgment Framework

Large Language Models (LLMs) as judges across various scenarios such as assessing model responses is becoming an increasingly accepted paradigm. However, existing judgment approaches often rely on trained judgers using fixed preference data, which tend to overlook diverse user preferences and struggle to adapt to real-world human-AI dialogue scenarios. To address these limitations, we propose SenseJudge, a customizable judgment framework driven by human preferences and SenseBench, a diverse and challenging instruction-following benchmark derived from real-world multi-turn interactions. We applied the automatic judgment framework and benchmark to two tasks: (1) LLMs as personalized judges, and (2) model ranking. We conducted extensive experiments, and the results demonstrate that the SenseJudge framework surpasses other judgment methods and models in the LLMs-as-personalized-judges task and achieves model ranking that aligns with real human sense. Additionally, we conducted analyses on position bias and consistency, alongside ablation studies, which affirmed the robustness of SenseJudge.

Rui Li, Junfeng Liu, Xiangwen Kong +2

State Key Laboratory for Multimedia Information Processing, School of Computer Science, Peking University · StepFun · Xi’an Jiaotong University

LLM Evaluation LLM-as-a-Judge Generative AI Evaluation