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

Aug 5, 2026cs.CL

Evidence Lock Before Commitment: A Frozen Interface Degrades LLM-as-Judge Evaluation

LLM judges are often asked to extract criteria and evidence before choosing between candidate answers. This workflow assumes that the intermediate record preserves the information needed for a later verdict. For reasoning-capable models, visible field order does not reveal internal decision order, so we test an observable alternative: persist the evidence in one call and make it the exclusive input to the next. Across 24,000 judgments over HelpSteer3, FeedbackQA, and CoVal, we compare standard pairwise judging, structured one-call judging, two-call evidence locking, and three-call pointwise locking with Claude Sonnet 4.5 and GPT-5. Evidence locking reduces agreement with released human preferences by 4 to 6 percentage points and increases answer-order inconsistency by 8 to 10 points relative to structured one-call judging. Pointwise locking is also harmful, while structured evidence elicitation remains close to standard judging. The result holds for both judges and all three datasets. Persisted evidence can support auditability, but it should not replace the source answers at decision time.

Divyansh Singh

University of Florida

Human Preference Evaluation LLM Evaluation LLM-as-a-Judge Pairwise Preference Evaluation

Aug 5, 2026cs.LG

Sharding Prevents LLM Oversight Failures and Adversarial Exploitation

Giving an LLM judge more compute does not necessarily make it check more requirements. When one call must return many verdicts, some decisions become weakly grounded in the evidence, even when that call receives the same token or tool budget as a panel of separate calls. Across expert-graded research replications, legal work, and clinical-trial assessments, agreement with experts falls as the number of verdicts per call grows. We identify sharding as the intervention that mitigates this failure in model-based oversight. Sharding partitions the requirements into smaller groups, assigns each group to a separate call, and aggregates the verdicts. Against a single call with the panel's full budget, sharding improves agreement while holding the model, evidence, total budget, and per-decision budget fixed. Overall, we find that a sharded weaker judge can outperform a more capable holistic judge and match that judge even when the latter receives the panel's full budget. Additionally, we find that sharding exhibits robustness against adversaries. A best-of-N adversary can hold the underlying work fixed, vary only its presentation, and increase an overloaded judge's acceptance of genuinely unmet criteria severalfold. Wherever sharding reduces baseline error, it removes this adversarial advantage, keeping over-acceptance low even as the adversary's search widens. Sharding does not address attacks that persuade the judge separately on each criterion rather than exploiting overload. In that setting, we find that debate-style opposition on top of sharding withstands such adaptive re-optimization.

Victor Akinwande, J. Zico Kolter, Aran Nayebi

School of Computer Science 2Neuroscience Institute Carnegie Mellon University

Scalable Oversight LLM Evaluation LLM-as-a-Judge Adversarial Attacks on LLMs Adversarial Robustness of Language Models

Aug 5, 2026cs.CL

The Evaluator Is Part of the Experiment: Measuring Open-Ended LLM Conformity

Prior work on LLM conformity largely measures discrete answer flips under verifiable labels. Open-ended revisions require a different measurement strategy because answer quality is graded, latent, and judged imperfectly. We introduce an experimental protocol implemented across a pooled main peer-condition corpus and separately constructed decomposition corpora, allowing us to separate ordinary re-answering, candidate-content exposure, a bundled peer-presentation residual, and directional judge sensitivity to visible peer context. Across four open-weight generators and three benchmarks, all-wrong peer input produces the lowest-quality revisions in every generator-dataset cell. Blind and informed ratings of identical answers also differ by evaluator: one judge shifts toward the peer-endorsed position, two shift away, one is approximately neutral, and GPT-4o and GPT-5.4-mini audits are likewise non-neutral. Finally, an anchor audit shows that terse correct anchors can be misread often enough to destabilize the latent scale unless calibration is checked explicitly. These results support four conclusions: flip rates are insufficient as a complete measure of open-ended conformity, wrong peers harm open-ended revision, evaluators are not neutral, and anchor calibration is necessary.

Alicia Guerra, Yibo Hu

Illinois Institute of Technology

LLM Evaluation LLM-as-a-Judge Bias in LLM-as-a-Judge

Aug 4, 2026cs.CL

TQLite: Multi-LLM Jury Guided Distillation for Real-time MQM Translation Quality Evaluation

Large language models (LLMs) have demonstrated impressive performance in MQM-based translation quality (TQ) evaluation, and recent advances in large reasoning models (LRMs) promise even greater improvements. However, both LLMs and LRMs are computationally expensive to deploy at scale, while small language models (SLMs)---though much more efficient---struggle with the complex reasoning required for evaluation tasks. In this work, we present an extensive empirical study benchmarking SLMs, LLMs, and LRMs across a wide range of TQ evaluation setups, providing a comprehensive view of the current landscape and establishing best practices. To address the scalability challenge, we introduce TQLite, a novel distillation framework that enables SLMs to approach the MQM evaluation performance of the best LRM-based evaluators. Our approach leverages a multi-LRM jury to generate high-quality synthetic training data via practical data curation techniques and aggregation of evaluation responses across a diverse panel of models. Our results demonstrate that SLMs trained via TQLite achieve strong MQM evaluation performance that far exceeds off-the-shelf evaluation capabilities of standard SLMs, offering a scalable and cost-effective alternative to LLM- and LRM-based evaluators.

Bhavin Jawade, Cameron R. Wolfe

Netflix

LLM Evaluation LLM-as-a-Judge Small Language Models Language Model Distillation MT Evaluation

Aug 3, 2026cs.CL

RADAR: Rubric-Aware Dependency and Redundancy Analysis for LLM-as-Judge Evaluation

Rubric-based LLM-as-judge pipelines often assume that evaluation criteria provide independent signals. In practice, however, criteria can be behaviorally coupled: improving one criterion may systematically change scores on another, distorting aggregate scores used in model-release or product-update decisions. We introduce RADAR, a lightweight preflight diagnostic framework for estimating such coupling before large-scale evaluation. Given a rubric, RADAR generates targeted synthetic probes, scores each probe on all criteria, and produces a directional coupling matrix that shows which criteria co-score and how. We validate RADAR on three industry-relevant evaluation settings: NVIDIA HelpSteer2, SumPubMed, and the Yale-Salesforce SummEval benchmark. Using only a small number of probes per criterion, RADAR recovers human inter-criterion correlation structure (Pearson r > 0.84) and provides practitioners with concrete audit signals about redundancy, hierarchy, and aggregation sensitivity before committing to large-scale judging.

Divyansh Singh, Reza Davari, Afra Mashhadi

Microsoft · University of Florida · University of Washington

LLM-as-a-Judge LLM Auditing Rubric-Based Evaluation

Aug 3, 2026cs.SD

Comparative Validation of GPT-4o-mini and Teacher Mean Scores for Automated Scoring of Music Analysis Responses: Single-Pass Deployment, Repeatability, and Strategy-Specific Bias

Scoring open-ended music analysis responses is time-consuming and requires nuanced judgments of harmonic knowledge and formal understanding. This study evaluates the validity and repeatability of GPT-4o-mini for rubric-based scoring of music analysis essays, using teacher mean scores as the benchmark. A dataset of 300 university-level student responses was scored by teachers on four dimensions: Harmony, Form, Reasoning, and Terminology. GPT-4o-mini scored the same responses using three prompting strategies: few-shot prompting with chain-of-thought reasoning (Fs+CoT), retrieval-augmented generation (RAG), and self-consistency based on five internal generations per administration (SC). Each strategy was administered three times with the model, prompt, rubric, and response held constant. Single-pass scores represented an operational scoring condition, whereas median aggregation across three runs was used to examine robustness. Agreement with teacher mean scores was evaluated using correlation, intraclass correlation, Krippendorff's alpha, quadratic weighted kappa, and scoring error indices. Fs+CoT showed the strongest agreement with teacher mean scores in both single-pass scoring and median aggregation. RAG showed systematic over-scoring, whereas SC produced highly repeatable scores but weaker individual-level agreement. Dimension-level analyses showed that scoring performance varied across rubric components, with Terminology generally showing weaker agreement than Reasoning. These findings indicate that GPT-4o-mini can generate stable scores for complex music analysis responses, but prompting strategies produce distinct scoring profiles. Operational use therefore requires strategy-specific calibration, dimension-level validation, and continued human oversight.

Baicheng Lin, Lingxi Jin, Kyung-Seok Min

Department of Education, Sejong University, South Korea · Department of Educational Technology, Ewha Womans University, South Korea

LLM-as-a-Judge Educational Technology Automated Essay Scoring Psychometric Validation Rubric-Based Evaluation

Aug 3, 2026cs.CL

Style Wins, Substance Loses: A Diagnosis of LLM-as-Judge in Idea Generation

However, whether these judges truly evaluate the scientific substance of ideas or are influenced by superficial stylistic presentation remains an open question. To address this question, we propose SciStyleBench, a unified three-component benchmark for diagnosing and mitigating stylistic bias in LLM-based idea evaluation: (i) First, SciStyleStage, a three-stage evaluation environment that applies controlled stylistic perturbations to fixed scientific content across three settings no context, fixed-domain context, and open-domain retrieval context, covering 600 scientific ideas and 15 style variants, with 9,000 evaluation instances per setting; (ii) Second, SciStyleMetrics, a set of quantitative measures, including Style Bias Index (SBI), Substance Recognition Rate (SRR), and Adversarial Win Rate (AWR), to characterize how stylistic variation affects scoring stability, substance discrimination, and ranking robustness; (iii) Third, SciStyleExtractor, a plug-and-play evaluation module that separates presentation style from scientific content by predicting style type and deviation before style-conditioned evaluation, enabling us to assess whether style awareness reduces stylistic bias. Experiments on SciStyleBench show that direct LLM judges remain sensitive to writing style and struggle to distinguish scientific substance. In contrast, SciStyleExtractor reduces SBI from 0.566 to 0.501 while increasing SRR and AWR from 0.504 and 0.554 to 0.759 and 0.899, respectively. These results suggest that robust idea evaluation requires invariance to stylistic variation without sacrificing sensitivity to scientific substance. Overall, SciStyleBench provides a systematic framework for identifying, quantifying, and mitigating stylistic bias in scientific idea evaluation.

Fengxian Ji, Yuke Li, Jingpu Yang +8

Wuhan University · Northeastern · Zhongguancun Academy +3

LLM Evaluation LLM-as-a-Judge

Aug 2, 2026cs.AI

Judging Is Not Enumerating: Silent Omissions in LLM-Authored Acceptable Sets

Language models are increasingly promoted from examinees to examiners: they write the test suites, answer keys, rubrics, and reward functions that define correctness for other systems. We measure the capability that role assumes and find it lacking under the protocol the role is usually deployed with, one-shot greedy authoring with no test-time reasoning. Across four reference constructions - two with complete finite truth, one with a hardened executable reference (HumanEval+/MBPP+), one with an explicitly incomplete lexical reference (WordNet) - models judge whether a candidate belongs far better than they author the set itself. On the incompleteness-proof algorithmic construction the gap is +0.34 to +0.29 F1 over a 24x parameter range and does not close; on executable code, models judging at F1 0.74-0.90 author suites admitting only 19-42% of oracle-correct solutions. A control locates the deficit: asked to emit the predicate rather than its extension, the same models reach F1 about 0.99. The failure is not missing knowledge or an inability to specify, but an inability to materialise the region a specification induces. The dominant error is omission, which resists audit: an over-inclusion is a token a reviewer can challenge, a missing member an absence whose discovery is the authoring problem itself. Models detect planted over-inclusions 6-7x more often than planted omissions, and a production deployment of 43,227 items fails omission-first at 10:1. Wired into RLVR, an authored key costs 1.9 points of accuracy against an exact oracle and 18.5 WordNet-relative (six paired seeds, p=0.031). Gating authored verifiers on a known-correct probe cuts false rejection from 58-92% to at most 5%, but keeps only 5-39% of suites. Repairing them instead, by rewriting each wrong expected value to what a reference execution returns, raises yield 3.3-10.6x across four author families.

Wenhui Chen, Jianlin Chen, Ziyao Lin +2

University of Macau · South China University of Technology · Independent Researcher

LLM Evaluation LLM-as-a-Judge LLM Auditing Automated Test Generation

Aug 1, 2026cs.IR

A Triple-Robustness Analysis of Retrieval-Augmented Generation for Multi-Hop Requirements Traceability

Reported verdicts on GraphRAG versus vector RAG disagree, and the evidence is typically tied to a single corpus, embedder, and judge -- and, we show, to where citation quality is measured. We present a triple-robustness analysis that holds a five-pipeline architecture matrix fixed and varies embedder (local e5-small vs. Azure text-embedding-3-small), corpus (DO-178C typed-edge requirements vs. Wikipedia paragraph chains via MuSiQue), and judge (paired GPT-5.4 x GPT-4.1 on both corpora), over 2x4,440 main-matrix runs, 600 cross-corpus runs, and over 5,000 faithfulness judgments. (C2a) GraphRAG's graph walk floods the context window at precision 0.12-0.23, but the synthesizer cites selectively at precision 0.48-0.65; scoring the retrieved set as the attribution set inverts the architecture ranking, which reconciles part of the disagreement in prior reports. (C1) Answer-level citation winners are corpus- and stratum-conditional but embedder-robust: GraphRAG ties vanilla on short-hop DO-178C queries and wins every MuSiQue stratum, while agentic pipelines lead only on 3+-hop requirements queries. (C2b) Faithfulness is corpus-conditional: on DO-178C it declines with hop distance (trend p<0.05 in three of four judge x embedder combinations); on Wikipedia chains neither judge shows a collapse. (C3) Single-judge LLM faithfulness is fragile to retrieval state: GPT-5.4's self-kappa across embedders is 0.137 (41% verdict change) against a same-day test-retest floor of 0.76, and re-judging frozen inputs eleven weeks later gives kappa <= 0.14 for both judges. A learned router on dense embeddings alone reaches macro-F1 0.86 on hop classification (C4). We argue that RAG architecture claims should be tested at this level of robustness -- including robustness to the citation-measurement point -- before they are trusted.

Meftun Akarsu, Burak Özdemir, Doğancan Büyükçolak +1

Turkish Aerospace Industries · Technical University of Munich

LLM-as-a-Judge Graph Retrieval-Augmented Generation Retrieval Robustness RAG Evaluation Multi-Hop Reasoning

Jul 31, 2026cs.AI

More Debate, Same Evidence: Structural Limits of Homogeneous Multi-Agent Groundedness

Large language model (LLM) judges are increasingly organized as multi-agent panels under the assumption that exchanging critiques improves judgment quality. We test this assumption for \emph{groundedness verification}, where a judge must determine whether a claim is supported by the supplied evidence. We evaluate a homogeneous three-agent panel on six public fact-verification and hallucination-detection benchmarks. Relative to a fixed single-agent reference, the panel's system-level accuracy difference ranges from

+8.5

to

-4.4

percentage points: two datasets show reliable gains, one shows a reliable loss, and three are statistically inconclusive. Because the reference and panel use different model variants, these differences characterize the complete systems rather than isolate a causal debate effect.

Yuelyu Ji

LLM Evaluation LLM Grounding LLM-as-a-Judge Multi-Agent Debate

Jul 30, 2026cs.CL

Beyond a Single Judge: The Evidence-Grounded, Social-Weighted Persona Panel for Generative UI Evaluation

Generative UI (GenUI) lets large language models synthesize a complete, renderable interface directly from a natural-language instruction, but evaluating the quality of what they generate remains an open problem. Human evaluation is costly and rater-variant, while LLM-as-a-judge is scalable but reflects only a single implicit viewpoint, unable to capture how different populations of real users actually perceive the same interface. We propose the Evidence-Grounded, Social-Weighted Persona Panel (ESPP), a three-stage GenUI evaluation method in which a panel of psychologically diverse, evidence-grounded personas independently rates a screenshot, exchanges opinions under a trait-derived, semantically-gated bounded-confidence mechanism, and is aggregated via Delphi-inspired social weighting into a single judgment. ESPP tracks human judgment substantially more closely than a naive single-pass judge, raising Pearson

r

from

0.716

to

0.922

, and a prompt-ensemble control recovers only about a third of this gap, isolating genuine persona and evidence grounding as the dominant source of improvement. Beyond this fidelity gain, retaining each panelist's individual rating further reveals that user subgroups agree on overall model rankings yet diverge sharply on specific rating dimensions, a structural disagreement a single homogeneous judge would systematically erase. The codes are available at https://github.com/Wuzheng02/ESPP.

Zheng Wu, Yibo Luo, Pu Zhang +2

School of Computer Science, Shanghai Jiao Tong University · 2ByteDance Inc

Human Preference Evaluation LLM-as-a-Judge Language Model Generation Evaluation Generative AI Evaluation

Jul 30, 2026cs.CL

(Towards) Scalable Reliable Automated Evaluation with Large Language Models

Evaluating the quality and relevance of textual outputs from Large Language Models (LLMs) remains challenging and resource-intensive. Existing automated metrics often fail to capture the complexity and variability inherent in LLM-generated outputs. Moreover, these metrics typically rely on explicit reference standards, limiting their use mostly to domains with objective benchmarks. This work introduces a novel evaluation framework designed to approximate expert-level assessments of LLM-generated content. The proposed method employs pairwise comparisons of outputs by multiple LLMs, reducing biases from individual models. An Elo rating system is used to generate stable and interpretable rankings. Adjustable agreement thresholds, from full unanimity to majority voting, allow flexible control over evaluation confidence and coverage. The method's effectiveness is demonstrated through evaluating competency profiles extracted from scientific abstracts. Preliminary results show that automatically derived rankings correlate well with expert judgments, significantly reducing the need for extensive human intervention. By offering a scalable, consistent, and domain-agnostic evaluation layer, the framework supports more efficient and reliable quality assessments of LLM outputs across diverse applications.

Bertil Braun, Martin Forell

KIT

LLM Evaluation Pairwise Comparison LLM-as-a-Judge LLM Reliability Automated Evaluation

Jul 30, 2026cs.CL

Rethinking LLM-Judged Helpfulness as a Pedagogy Signal: A Pre-Registered Audit Across Tutor Models

LLM tutoring poses a measurement problem: can a general-purpose helpfulness rubric distinguish direct answer-giving from pedagogical guidance? We audit this signal in a pre-registered study. Within each of three tutor bases, we compare conversational and pedagogical policies instantiated with the same underlying model and paired with one fixed weak simulated student. Deterministic detectors measure answer leakage and next-turn independent work. Claude Opus 4.8 is the frozen, condition-blind primary judge. After the Opus scores were fixed, GPT-5.6 Sol was prospectively specified for a post hoc robustness audit of the same 1,179 confirmatory answer-phase tutor turns under the frozen helpfulness and pedagogy rubrics. On the primary base under Opus, the policies do not differ significantly in helpfulness but are perfectly rank-separated under the pedagogy rubric (Cliff's

|δ|{=}0.10

vs.

1.0

). Across the two judges, pedagogy contrasts retain their direction where detected, whereas the helpfulness ordering is judge-contingent, reversing between judges on two of three bases. In an Opus-only ablation, seven primary-base policies span

2.3

points in mean judged pedagogy within a

0.25

-point band of mean judged helpfulness. Separately, answer-revealing turns are followed by less independent student work on every base, a result that is judge-invariant by construction. In this controlled setting, general-purpose helpfulness is not a reliable pedagogy signal. Tutor evaluation should pair pedagogy-targeted rubrics with deterministic process measures.

Shuyi Fan, Boyuan Deng, Mengyu Xu +4

Columbia University, USA · Johns Hopkins University, USA · The University of Chicago, USA +4

LLM Evaluation LLM-as-a-Judge Intelligent Tutoring Systems Generative AI in Education

Jul 30, 2026cs.AI

Share the Judge, Learn the Deferral: Where Specialization Helps LLM Evaluation

Agentic systems have widened the gap between producing candidate outputs and reviewing them. This paper asks a practical architectural question: should domain specialization be built into an evaluator's weights, or into the rule that decides when its judgment can be trusted? We study 99,952 public, rubric-conditioned examples. Supplying the correct rubric improves locked-test accuracy by 2.11 points over a response-only control; replacing it with an unrelated rubric costs 2.66 points. Dividing the same training corpus among eight criterion-family LoRA judges, however, loses 10.05 points and cuts audited coverage at a 5% risk target from 24.44% to 5.43%. Matching the bank's stored capacity with one rank-64 adapter does not reproduce this loss. Nor is the result explained by learning rate or optimizer steps. Initializing the family adapters from a shared, trained judge recovers test accuracy to 76.85%, 19.94 points above scratch training at the same learning rate (95% interval 18.88-21.02). The result changes when specialization governs deferral rather than judgment. On RewardBench 2, learned correctness heads route examples through a 0.6B-4B-8B cascade without changing any reward score. Across 20 locked repartitions, the cascade attains 89.40% accuracy, compared with 84.75% for 8B alone, at 0.415 normalized parameter compute. Every run passes an exact one-sided 95% risk audit; margin-based rules remain near 84.8% accuracy while using at least 0.94 compute. These results suggest a qualified design rule: share the learning of judgment until there is enough data to justify a split, and place domain-specific adaptation in an audited release boundary.

Weining Zhang

Cheung Kong Graduate School of Business

LLM Evaluation LLM-as-a-Judge Selective Prediction Rubric-Based Evaluation

Jul 25, 2026cs.AI

SeekJudge: A Practical Reward Framework for Reinforcement Learning in Computer-Use Agents

Deciding whether a trajectory actually fulfills its instruction governs how we measure computer-use agents on long-horizon graphical-user-interface tasks and how we train them with reinforcement learning. This judgment has long relied on rule-based evaluation, which struggles to align with human intention and goes stale when an app updates or its online content drifts. Existing model-based judges attempt to address these problems but still leave a performance gap to the rule-based evaluation. We propose the \textbf{SeekJudge} framework, in which four role-specialized agents, a Condense, a Ground, a Seek and an Analyze agent, reach a verdict through a Seek--Analyze loop over the trajectory. A seed-calibrated distillation pipeline trains one specialized

9

B model to serve as the shared backbone for all four agents. Measured by downstream success rate on held-out RL test goals, SeekJudge is the first practical model-based reward to match or surpass native rule-based supervision in online RL. Beyond accuracy, SeekJudge provides step-level judgments, runs far cheaper than a closed-source large model, and keeps a small per-call context that scales to much longer trajectories. We further contribute a general architectural improvement to the reward server that speeds up judging in RL. Together these make model-based reward a practical drop-in for rule-based supervision in CUA reinforcement learning.

Yang Wan, Zhenhao Zhang, Jierui Wang +1

1Zhejiang University · University of Rochester

Reward Modeling LLM-as-a-Judge RL for GUI Agents Computer-Use Agents Long-Horizon Agent Evaluation Reward Model Evaluation+1

Jul 22, 2026cs.CL

Reference-Free Evaluation of Reasoning in Open-Ended Question Answering

AI-generated answers in high-stakes domains are often fluent but difficult to verify, especially when they contain multi-step reasoning rather than a single final answer. We propose a reasoning-based, reference-free framework for auditing LLM-generated outputs. The method decomposes a generated reasoning trace into segments, labels local premise-target relations using Natural Language Inference (NLI), and organizes these relations into a hypergraph. A deterministic backward AND-OR search then assigns segment-level audit labels that indicate how each segment is grounded within the generated response. We evaluate the framework in two settings: deductive mathematical reasoning with Hard2Verify, and open-ended medical reasoning with UroReason, a new physician-annotated benchmark of LLM reasoning traces from real clinical cases. Across these settings, our NLI-hypergraph audit provides a more reliable reference-free evaluation signal than direct LLM-as-judge baselines. In the clinical setting, state-of-the-art LLM judges often fail to identify problematic reasoning segments, over-accepting fluent but weakly grounded responses. Our results show that QA evaluation should account for how inferential relations compose across a reasoning trace, rather than relying only on final answers or LLMs as verifiers. UroReason will be made available through an API, and our code will be released as open source.

Guneet Singh Kohli, Yuxiang Zhou, Michael Sejr Schlichtkrull +2

Queen Mary University of London · Temple University

LLM-as-a-Judge Reasoning Evaluation LLM Auditing Clinical Language Model Evaluation

Jul 21, 2026cs.CL

Two-Level Meta-Rubrics for Evaluating Open-Ended Generation: GAMUT, a Benchmark for Factual Completeness

Rubric-based evaluation of open-ended generation faces a fundamental tension between expressiveness and reliability. Authoring a faithful rubric requires expressing the structure of the space of good answers: open-ended sets of acceptable options, ordered processes, and the relative importance of facts. Grading with the rubric requires a judge to score consistently, and judges are far more reliable on flat, binary checks than on rich structure. We resolve this tension with a two-level meta-rubric framework. A structured meta-rubric captures the grading criteria at authoring time, and fixed mechanical rules compile it into a flat checklist of binary, machine-gradable checks that an LLM judge scores reliably at evaluation time. We instantiate the framework as Gamut (Grounded Assessment of Multimodal Factuality), a benchmark for factual completeness in long-form generation. Gamut comprises 1,813 questions grounded in real wearable imagery across 10 diverse domains, each paired with an evidence-backed rubric verified by expert human annotators. Evaluating 14 frontier and open-weight models, we find Gamut genuinely challenging (best score 58.7% from Gemini 3.1 Pro), highly discriminative, and robust to the choice of judge.

Xilun Chen, Zhaleh Feizollahi, Ross Goodwin +5

Meta AI

Open-Ended Generation LLM-as-a-Judge Language Model Generation Evaluation Rubric-Based Evaluation Automated Evaluation Multimodal Model Evaluation

Jul 21, 2026cs.LG

SFGA: A Statistics-First Gating Architecture with Adjudicative Escalation for Trustworthy SFT Data Procurement

Procuring supervised fine-tuning (SFT) data forces a buyer to decide, before any downstream training, whether a candidate corpus is worth acquiring. We present \sys{}, a statistics-first gating architecture that treats procurement as a cost-aware routing problem over three intrinsic quality axes -- diversity, utility, and redundancy. Cheap blind measurements are summarised into per-axis estimates with confidence intervals; a gate accepts a decision only when intervals are tight, sample sizes are adequate, and the axes agree, otherwise it escalates the case to an adjudicative debate between a buy-advocate and a reject-advocate judge, resolved by a presiding verdict. On a controlled benchmark of 12 datasets (

2{\times}3{\times}2

grid over the three axes) with 5 seeds, the gate reaches 0.90 accuracy and 0.83

F_1

at $0.017 per unit, sitting between an always-verify baseline (0.75) and an oracle upper bound (0.98) while spending less than always-escalate ($0.020). We further report honest negative diagnostics of the debate path: a con-side win rate of 0.80 (

p\approx3{\times}10^{-6}

) and a 52% position-flip rate under advocate swapping expose negativity and positional biases that a naive LLM-judge would hide. We frame the injected-knob evaluation explicitly as a controlled synthetic benchmark for measurement fidelity and routing calibration, and delimit external validity as future work.

Arther Tian, Alex Ding, Simon Wu +1

DGrid AI

Supervised Fine-Tuning LLM-as-a-Judge Selective Prediction

Jul 21, 2026cs.AI

Evaluating medical AI under missing information: same-provider judges and human raters change apparent safety

Readiness stress-testing of medical AI has focused on closed-ended and multimodal benchmarks. We extend it to open-ended clinical conversation under missing information, where safe behavior means recognizing absent information and qualifying, clarifying, or not over-committing - and where the evaluator becomes part of the measurement. We stress-test four models - three flagships (Claude Opus 4.8, GPT-5.5, Grok 4.3) and one mid-tier model (Gemini 3.5 Flash) - by deleting the latter half of the final user turn in HealthBench conversations, grading responses with a four-provider LLM-judge panel and a blinded clinician-anchored reference. Two evaluator-facing results are robust. First, judge choice materially changes apparent safety: inter-judge agreement is only moderate (Fleiss' kappa = 0.65), and after adjusting for each judge's general leniency (vote-level logistic regression), a positive same-provider association remains (exact permutation p = 0.04; GPT-5.5 ~ +0.10 on the probability scale) - large enough to change which model appears to over-commit least once its own-provider judge is excluded. Second, LLM judges are more permissive than clinicians on a blinded 50-item subsample: all four are significantly more lenient than the stricter independent clinician (crediting appropriate uncertainty on 66-84% of items vs 52%), and three of four than the author-influenced consensus (Grok directional only; judge-vs-consensus kappa = 0.20-0.43). On the author-audited clinical-underdetermined subset the permissiveness gap widened and the point-estimate model ordering held. A closed-ended MedQA anchor confirms accuracy is high and option-order effects are within a +/-5-point equivalence region for three of four models, so the safety gap is about calibration, not knowledge. We release the harness, prompts, per-item outputs, judge panel, perturbation audit, and human-annotation protocol.

Koyar Afrasyab

Kinvectum AB, Sweden

LLM-as-a-Judge Language Model Safety Evaluation Clinical Language Model Evaluation Uncertainty Calibration

Jul 20, 2026cs.HC

EduPanel: A Three-Agent LLM Judge for Teaching Videos -- Reliability, Complementarity, and Human Trust Calibration

Teaching videos are becoming a major medium for education, creating a growing need for scalable evaluation of their pedagogical quality. Existing automatic judges do not fully address this setting because teaching quality depends on multimodal evidence and should be evaluated with respect to the intended learner rather than as a universal property. We present EduPanel, a rubric-grounded, learner-conditioned LLM judge that decomposes evaluation across specialized agents to produce interpretable assessments for different aspects of teaching quality. Across expert studies, architecture ablations, and learner-persona analyses, EduPanel achieves reliability comparable to a median human expert. In expert evaluation, its feedback improves scoring accuracy (MAE 0.87 to 0.73), while experts remain able to detect unreliable outputs (AUC = 0.77) instead of accepting them blindly. These results suggest that EduPanel can serve as effective assistants for educational evaluation rather than replacements for human experts.

Jia-Kai Dong, Yi-Cheng Lin, Hung-yi Lee

National Taiwan University · NTU Artificial Intelligence Center of Research Excellence

LLM-as-a-Judge Educational Assessment Human-in-the-Loop Evaluation Multi-Agent Systems Rubric-Based Evaluation

Jul 20, 2026cs.AI

Judge-dependent safety gains and model-specific helpfulness costs of evidence-sufficiency prompting in clinical LLMs

Background: LLM judges increasingly score whether clinical language models give overconfident answers under incomplete evidence, yet whether a measured "safety gain" reflects real behavior change or the judge's calibration is unresolved. Using a structured evidence-sufficiency prompt as a test case, we asked whether it reduces unsafe overconfident answers, how far that effect depends on the scoring judge, and what it costs in helpfulness. Methods: In a retrospective public-data benchmark (Real-POCQi, HealthBench, MedRBench), four models (GPT-5.5, Claude Opus 4.8, Gemini 3.5 Flash, Grok 4.3) answered a fully paired common panel (1,200 cells) with a standard prompt and the wrapper. The pre-specified endpoint was the paired reduction in unsafe overconfidence scored by the primary judge (GPT-5.4-nano); secondary analyses added a different-family judge (Claude Sonnet 5), a correctness judge, matched scaffold controls, and a blinded three-clinician review. Results: Unsafe overconfidence fell from 49.3% to 24.7%, a paired reduction of 24.7 points (95% CI 21.8-27.7; p<0.001), robust in direction across models and paraphrases. Magnitude was judge-dependent: Sonnet agreed on direction but nearly halved the effect (+13.1 points), with one-directional disagreement. Blinded clinicians characterized the primary judge as a high-sensitivity (1.00), low-specificity (0.55) screen, not a calibrated rate. The gain carried a model-specific helpfulness cost (correct diagnosis 80.3% to 50.3%): near-free for GPT-5.5, near-total for Gemini (-58 points). Matched scaffold controls showed genuine behavior change, not judge circularity. Conclusions: LLM-judged clinical safety effects should be reported as directional and relative, anchored to human review and evaluated jointly with helpfulness, not as calibrated absolute rates. This does not establish clinical deployment readiness.

Koyar Afrasyab

Kinvectum AB, Sweden

LLM-as-a-Judge LLM Prompting LLM Safety LLM Safety Evaluation Clinical Language Model Evaluation

Jul 19, 2026cs.CL

A Consensus-Based Framework for Relative Preference Evaluation of Large Language Models

Traditional benchmarks for LLMs primarily rely on static datasets and objective scoring metrics, which often fail to capture differences in response quality when multiple answers are acceptable. In such settings, correctness alone is insufficient to distinguish between responses that vary in clarity, completeness, and usefulness. This paper introduces a consensus-based evaluation framework that measures relative preference among model-generated responses rather than absolute correctness. Instead of evaluating outputs against a fixed ground truth, we assess how a panel of diverse LLMs ranks anonymized candidate responses to the same prompt. This approach treats aggregate inter-model agreement as a proxy for perceived response quality under blind conditions. We conduct a controlled study using five state-of-the-art LLMs across multiple domains, including programming, general knowledge, safety, logical reasoning, and mathematics. Each model generates responses and independently ranks peer outputs through a structured voting process. Scores are aggregated into a Relative Intelligence Index (RII), representing how frequently a model's responses are preferred by other models. Our findings reveal consistent preference patterns across domains, with certain models more frequently ranked highly by their peers. However, we emphasize that these results reflect inter-model preference alignment rather than objective correctness or human judgment. This framework provides a scalable, model-driven method for comparative evaluation, offering an alternative perspective on response quality in scenarios where multiple valid answers exist. While not directly aligned with human evaluation, prior work suggests that aggregated model preferences can partially correlate with human judgments, motivating this as a proxy signal.

Mohtashim Khan

LLM Evaluation LLM-as-a-Judge

Jul 16, 2026econ.GN

Does Multi-Agent Debate Improve AI Feedback on Research Papers?

Probably not, at least for meta-analyses in economics. In a pre-registered, identity-masked, within-paper experiment, the authors of 44 meta-analyses ranked three AI reports on their own paper by usefulness for improving it: a single pass by a frontier model against two multi-agent debate tools we built and expected to win. All reports were held to a common length and template. The authors preferred the single pass, by 0.66 rank points over mad-research (95% CI 0.32 to 1.00) and 0.57 over paper-workshop (0.16 to 0.95), though paper-workshop spent roughly thirty times the tokens. Authors who recalled their journal referee report usually placed it first and never last; in a separate exercise, three AI judges almost always placed the real journal referee report last. Among the three AI reports, Gemini (the judge whose model family wrote none of the reports) would have ranked paper-workshop first in the authors' place, reversing the single-pass preference. The reversal warns against substituting an AI judge for the author. We measure perceived usefulness for finished papers; whether AI should referee papers is a separate question.

Tomas Havranek, Zuzana Irsova

Charles University, Prague · Centre for Economic Policy Research, London · Meta-Research Innovation Center at Stanford

LLM-as-a-Judge Multi-Agent Debate Human-AI Decision Making

Jul 15, 2026cs.CL

The Test Oracle Problem in Synthetic LLM-as-Judge Corpora: Disappearance, Distortion and a Validation Protocol

Studies of bias in LLM-as-judge systems typically build synthetic corpora by prompting an LLM to generate a hallucinated answer to pair with a factual one, then presenting both to a judge. We report a case in which this generation step silently failed, and use it to argue that the failure mode is structural rather than incidental. In a multilingual (Turkish/English) faithfulness-judgment corpus, a decoding-budget parameter shared between judging and generation calls truncated one producer's hallucinated answers to a few words. The resulting items produced a large, statistically robust effect: a 32-point cross-lingual collapse in one judge's selection accuracy, replicated from N=50 to N=500, explained by a three-layer mechanistic account, and confirmed by a controlled producer-swap experiment, none of which was real. The effect vanished to ceiling once the shared parameter was corrected, and only manual reading of the raw generations, not any aggregate statistical check, exposed the fault. A second measured bias (Markdown-formatting preference) was not fabricated but distorted by the same fault, its magnitude and in one case its sign shifting with stimulus length, a mode aggregate metrics cannot distinguish from the first. We frame the underlying vulnerability using the test oracle problem: corpora whose negative examples are LLM-generated carry no mechanical way to verify item integrity, while corpora built by deterministic perturbation of a gold answer carry an item-level oracle for free. A positive control supports this claim directly: an analogous fault injected into a minimal perturbation-based corpus is caught with 100% accuracy by a zero-cost, zero-human gold-to-negative string comparison. We close with a validation protocol, derived from our own case, for analysts working in the oracle-less regime that we argue describes most contemporary multilingual LLM-as-judge corpora.

Serkan Ballı

Department of Software Engineering, Mehmet Akif Ersoy University, Burdur, Türkiye · Department of Software Engineering, Mehmet Akif Ersoy University, Türkiye

Multilingual Language Model Evaluation LLM-as-a-Judge Synthetic Benchmark Generation Language Model Bias Evaluation Bias in LLM-as-a-Judge

Jul 15, 2026cs.CL

Evaluation Ability Does Not Imply Optimization Utility: LLM-as-a-Judge Signals in Closed-Loop Table Recognition

LLM-as-a-judge is widely used to provide feedback and selection signals in closedloop regeneration, but this use remains insufficiently validated. We study it in table recognition, where deterministic TEDS evaluation provides a controlled testbed, using FinTabNet and OmniDocBench. Three findings emerge. First, judge signals were weak on both datasets: scores frequently tied, rankings were not reproducible, and the only selection policy that beat random on both datasets depended on an earliest-iteration tie rule, so its advantage cannot be attributed to the judge scores alone. Iteration produced better candidates, but the judge failed to recover them. Second, severe losses occurred even without specific judge feedback. A structurepreserving instruction significantly reduced the severe-loss rate on FinTabNet and was directionally consistent on OmniDocBench. The contrasts support target-preservation failure under unconstrained regeneration as a proximate mechanism of the observed severe losses. Third, the structure-preservation constraint reduced the severe-loss tail but produced no improvement. In an exploratory 2x2 analysis, the same protection was not stably observed when judge feedback was retained. These results do not dispute the value of LLMs as evaluators. Instead, they show that evaluation ability does not imply optimization utility. Iterative refinement requires, at minimum, a verification signal that deterministically detects structural change, rather than judge scores alone.

Donghwan Kim

Aidentyx Inc., San Jose, CA, USA

Table Structure Recognition LLM-as-a-Judge Iterative Refinement

Jul 14, 2026cs.AI

Rating the Raters: Rasch Measurement Theory for LLM Evaluation

LLMs now sit on every side of evaluation: as examinees scored on benchmarks, judges of other models' outputs, and raters of human-generated content. Each paradigm can be viewed as a measurement problem, where a latent property of an object is probed with items from an instrument (e.g., benchmark) by judges or raters. Standard evaluation practices often neglect the contributions of each core component to the end result, limiting our understanding of what is being measured. Rasch measurement theory (RMT) is well-suited to this problem. RMT decomposes ordinal ratings into separable facets on a common scale. It further provides a battery of diagnostics that can identify miscalibrated measurements and rater biases. We present a case study of RMT applied to the LLM-as-rater paradigm using the Measuring Hate Speech corpus, whose construct was itself built under RMT. We fit a series of many-facet Rasch models to annotations from nine LLMs spanning families and capability levels. Our analyses show that LLMs systematically differ from human raters in severity, item-level calibration, question-order robustness, target-identity sensitivity, and rating scale use, all of which standard evaluation practice would largely obscure. Overall, we argue that RMT belongs in the toolkit for evaluating LLM-as-examinee, -judge, and -rater paradigms.

Pratik S. Sachdeva, Nathan Boudol

University of California, Berkeley · Grenoble INP

LLM Evaluation LLM-as-a-Judge

Jul 14, 2026cs.CL

LLM Judges Can Be Too Generous When There Is No Reference Answer

LLM judges are increasingly being used to evaluate open-ended model responses, often in no-reference settings where a ground-truth answer is unavailable. However, can they reliably assess in such evaluation setups? We explore this question in this paper through a two stage pipeline with a) calibration experiments that assess the judge model's knowledge of the task it is evaluating, and b) sensitivity experiments that assess how the judge model's performance is impacted by the presence and positioning of the reference answer in the prompt. Across experiments covering three languages, we show that the judge models we evaluated tend to over-credit incorrect answers in the absence of a reference answer, and adding reference answer information to the prompt flips the judge model's correct/incorrect decisions by as much as 85% in some experimental settings. Comparison with a subset of human annotations shows that these reference-driven changes generally align with human judgments. Our results emphasize the need for calibrating the LLM judges with a sample with reference-aware evaluation before using them in reference-free setups reliably, and our methodology provides a blueprint for researchers and practitioners in doing such calibration of LLM judges for other tasks.

Chalamalasetti Kranti, Sowmya Vajjala

Computational Linguistics, Department of Linguistics, University of Potsdam, Germany · National Research Council, Ottawa, Canada

LLM Evaluation LLM-as-a-Judge Language Model Calibration Reference-Free Evaluation Bias in LLM-as-a-Judge

Jul 14, 2026cs.CL

FinResearchBench II: A Deep Research Benchmark with Consensus-Derived Gold Rubrics for Distinguishing Financial Report Quality

Deep research agents are increasingly used to produce long-form financial reports, yet large-scale evaluation remains bottlenecked by the need for human experts to define and execute high-quality rubrics. We address this problem by proposing a scalable pipeline for generating high-quality rubrics without human experts in the final loop. We build a financial deep research benchmark from 104 real-world user queries and automatically synthesize 14,450 query-specific candidate rubrics from model-generated reports. To justify removing human experts from rubric execution, we compare rubric judgments from three human experts with those from a three-LLM judge panel on a sampled subset, and show that LLM-based evaluation is sufficiently consistent with human evaluation to replace it for large-scale rubric screening, including 98.67% label-level agreement on jointly unanimous items. We then derive consensus-derived gold rubrics through two filters: a strict consistency filter, which keeps a rubric only if the three LLM judges unanimously agree on every report under the same query, and a distinguishability filter, which keeps a rubric only if it assigns at least one majority-yes and at least one majority-no label across the evaluated systems. This process retains 3,687 consistency-passed rubrics, of which 2,600 remain distinguishable and form the final set of consensus-derived gold rubrics. Using this final rubric set, we obtain clearly differentiated rankings across 10 deep research systems, with item-level pass rates ranging from 58.58% to 22.23%. More broadly, because the pipeline removes human-expert execution from rubric generation and evaluation, it is naturally scalable for benchmark evaluation, automatic system comparison, and future studies of evaluation-driven system improvement.

Beidi Luan, Rui Sun, Sinuo Wang +5

1StepFun · University of Adelaide · 2FinStep +1

LLM-as-a-Judge Rubric-Based Evaluation Rubric-Guided Generation Finance

Jul 13, 2026cs.LG

Inside the Unfair Judge: A Mechanistic Interpretability Account of LLM-as-Judge Bias

Existing studies of LLM-as-judge scoring bias work predominantly at the input-output level: they perturb inputs, measure score deltas, and propose prompt-level mitigations. We argue that the same biases admit a representation-level account in the judge's hidden state, complementary to the input-output view and operationally useful in ways it does not afford. We report three findings, across seven judges, seven bias types, and nine benchmarks. Geometry: baseline judging inputs occupy a tight activation manifold while biased inputs are displaced along a low-dimensional, type-specific subspace that sharpens with depth and is recovered consistently by three families of estimators. Causal control: steering hidden states along this subspace drives scoring in both directions, forward shifts reproducing biased scoring on clean inputs and reverse shifts restoring baseline scoring on biased ones, while matched-norm random directions produce shifts an order of magnitude smaller. Operational: a simple linear projection onto the same bias-direction features anticipates judge failures on three entirely unseen benchmarks, substantially outperforming text-based alternatives. Reading bias as activation geometry, rather than as input-output noise, unifies geometric structure, causal control, and operational prediction within a single framework. The project page is available at https://xzx34.github.io/unfair-judge/

Zixiang Xu, Sixian Li, Huaxing Liu +4

AMAP, Alibaba Group · Mohamed bin Zayed University of Artificial Intelligence · University of Southern California +1

LLM-as-a-Judge LLM Interpretability Language Model Bias Evaluation Bias in LLM-as-a-Judge Representation Geometry in Language Models

Jul 12, 2026cs.CL

Knowledge Distillation for Automated AI Tutor Evaluation

The rapid integration of Large Language Models (LLMs) into K-12 and higher education has outpaced the development of reliable methods for evaluating their pedagogical quality. As the research community starts to explore the space of automating evaluation of AI tutors, we introduce FATE (FLC AI Tutor Evaluator), a specialized 8B-parameter language model designed to evaluate AI tutors. Aligned with the four core evaluation tracks from the BEA 2025 Shared Task, our model assesses pedagogical ability across Mistake Identification, Mistake Location, Guidance, and Actionability. Because pedagogical evaluation is a specialized task with limited labeled data, we leverage knowledge distillation from a frontier LLM to generate additional supervision, yielding absolute performance gains up to 22.63 percentage points. Finally, we demonstrate FATE's utility as an automated evaluator by benchmarking instructional responses generated by popular commercial models, including ChatGPT, Claude, Gemini, and DeepSeek. On average, we have found that Gemini 2.5 Flash perfomed best (82.88%), then ChatGPT 5.5 Instant (80.75%), followed by DeepSeek V4 Flash (80.13%) and Claude Sonnet 4.6 (74.00%).

Tahmid Al Hannan, Diego Garcia, Alex Njoroge +2

Folsom Lake College Folsom, CA, USA

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