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

Sep 29, 2026cs.SE

CRJudgeBench: Can AI Detect Plausible but Invalid Code Reviews?

Large language models can generate plausible code-review comments, but such comments may contain technically incorrect claims that mislead developers. We study technical trustworthiness judgment: determining whether a review comment's core technical claims are correct and applicable to the reviewed code in its repository context. Existing code-review benchmarks primarily evaluate review generation, issue discovery, or general comment quality, but do not directly assess whether an agent can determine the technical trustworthy of an individual review comment. To fill this gap, we introduce CRJudgeBench, a benchmark of 1199 instances constructed from real pull requests and expert-verified perturbations, covering both trustworthy and plausible but untrustworthy comments. We further present Sentinel, a repository-grounded agentic judge that actively gathers code evidence to verify review comments before making judgments. Starting from Qwen3-Coder-30B-A3B-Instruct, Sentinel is trained on the CRJudgeBench training split through iterative action-level learning from a privileged teacher. On the 359-instance CRJudgeBench test set, Sentinel achieves 76.60% accuracy, outperforming GLM-5.3 by 6.13 percentage points and its base model by 19.78 points. These results show that even state-of-the-art general-purpose LLMs struggle to identify untrustworthy comments, while iterative action-level learning substantially improves the accuracy of repository-grounded trustworthiness judgments. Our dataset is available at https://huggingface.co/datasets/dcloud347/CRJudgeBenchmark

Yue Pan, Jiawei Li, Ziyuan Zhang +2

University College London · Amazon · Zhejiang University

LLM Evaluation LLM-as-a-Judge AI Agent Evaluation Automated Code Review

Sep 29, 2026cs.AI

From Judgment Quality to Downstream Utility: Rethinking LLM-as-a-Judge for Open-Ended Tasks

LLM-as-a-Judge is increasingly used to evaluate policy responses on open-ended tasks that lack ground-truth answers. Existing work often directly converts the resulting judgments into reward signals for policy training, paying limited attention to intrinsic judgment quality and largely restricting the use of Judges to training-time supervision. We systematically investigate judgment quality and downstream utility by examining both how judgments are elicited and how they are used. For judgment elicitation, we vary the Judge protocol along three dimensions: verdict granularity, critique usage, and evaluation batching. For judgment usage, beyond policy training, we extend Judge to test-time inference through Best-of-N selection, Judge-guided revision, and beam search. We find that, (i) Surprisingly, judgment quality and downstream utility do not always align. (ii) Judge protocol design substantially affects both intrinsic judgment quality and downstream utility. (iii) Judge guidance effectively converts test-time compute into performance gains, with benefits varying across inference strategies. Our results call for a multifaceted evaluation of LLM Judges on open-ended tasks, encompassing intrinsic judgment quality, and downstream utility.

Zheng Zhang, Lufei Li, Xinyue Tan +4

School of Information Science and Technology, ShanghaiTech University · State Key Laboratory of General Artificial Intelligence, BIGAI

LLM Evaluation LLM-as-a-Judge

Sep 29, 2026eess.AS

Louder, Longer, Livelier: Acoustic Shortcuts and Underspecified Rationales in Speech LLM Judges

LLM-as-a-judge is widely used for evaluating text, but extending this paradigm to speech requires models to interpret acoustic as well as linguistic evidence. This introduces a modality-specific risk: a speech judge may treat a perceptually salient cue as evidence of quality even when that cue is irrelevant to the target criterion or receives more weight than human listeners give it. We call this behavior an acoustic shortcut. To study it, we audit six speech LLM judges using controlled manipulations of intensity, content richness, and emotional delivery. We evaluate both pointwise scoring and pairwise comparison, using human preference calibration to interpret the results. The judges consistently reward louder audio, prefer content-rich speech more strongly than human listeners do, and map emotional delivery into quality preferences. These effects are most visible in pairwise comparison, while pointwise scores often obscure them. More concerningly, the accompanying rationales rarely identify the acoustic cue that changes a judgment and instead repeatedly rely on a limited vocabulary, leaving them acoustically underspecified. Together, these findings show that reliable speech judges must both resist acoustic shortcuts and ground their rationales in the acoustic evidence behind their decisions. To support reproducibility and future audits, we also release SpeechJudgeAudit, the controlled stimuli and evaluation tools used in this study.

Mingyue Huo, Shivam Mehta, Bhavin Jawade +2

University of Illinois Urbana-Champaign · Netflix

LLM-as-a-Judge Speech Quality Assessment Speech Language Models Shortcut Learning Language Model Auditing

Sep 29, 2026cs.AI

JudgeProfile: Understanding and Steering Subjectivity in LLM Judges

LLM judges are inherently subjective, often favoring different responses in pairwise comparison when neither option is objectively wrong. To study this subjectivity, we introduce JudgeProfile, a framework that dissects LLM evaluation into perception (how a judge compares two responses across specific attributes like clarity, correctness, and detail) and prioritization (how much each attribute influences the final choice). We curate SubjectiveSet, a dataset of 50,013 response pairs from 17 public data sources, evaluated by 21 LLM judges across 87 attributes. We find a hidden consensus in perception: judges frequently agree on attribute judgments even when their overall choices diverge. Building on this separation, we first characterize each judge's prioritization using attribute weights estimated from its own overall choices. These weights differ across judges even when estimated from the same attribute judgments. We then learn new weights from reference labels to adapt their decisions to a target evaluation standard. Reweighting perceived attributes improves average held-out agreement with reference labels from 66.48% to 71.97%, outperforming fine-tuning and rubric prompting. Our findings show that understanding and steering the subjectivity of LLM judges requires attention not only to what they perceive, but also to how they prioritize it.

Qi Cao, Kangning Liu, Xuan Kan +10

Meta · University of California San Diego · The University of Hong Kong +1

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

Sep 28, 2026cs.AI

The Argument and the Letterhead: Source-Position Coherence in AI Evaluation

An argument can be surprising coming from a particular speaker without being a bad argument. Do AI evaluators keep these judgments apart? Two preregistered descriptive studies and a later Jev supplement collected 2,976 usable evaluations of six fixed texts about US AI policy, Germany's debt brake and Swiss nuclear energy. Each text was presented under several source attributions. The key comparison asks whether the gap between two sources changes when the argument changes. On Sol, for example, a national-security argument received mean ratings of 0.359 under CODEPINK and 0.639 under College Republicans; a civil-rights argument received 0.742 and 0.721. A constant preference for one source cannot explain that pattern. Related interactions appeared across topics and recent model configurations, including those with reasoning enabled, while several comparisons yielded small effects. The later European Jev supplement yielded five interactions below the adopted absolute reference of 0.05; its distinct rubric and interrupted collection limit comparison with the chat systems. Some written evaluations explicitly invoked a mismatch between a source and its attributed position. Taken together, the numerical and verbal evidence supports source-position coherence as a plausible explanation, alongside competing accounts involving credibility, authenticity and interpretation of the task. The paper develops this inference through controlled comparisons, reports conditional post hoc p-values in an appendix, and documents the human decisions and delegated checks behind an AI-conducted study.

Michele Loi

LLM-as-a-Judge Political Bias in Language Models Language Model Bias Evaluation

Sep 28, 2026cs.CL

Certified Selective Automation of LLM Agent Evaluation

Evaluating LLM agents still ends with a human reading trajectories, because automatic judges carry no guarantee on how often they are wrong. We ask the operational question: what fraction of agent evaluation can a judge take over, with a certificate that the error rate among auto-decided trajectories stays below a budget alpha? Agent corpora resist the standard answer: many agents attempt the same tasks, so trajectories arrive in correlated clusters, and the i.i.d. certificates of existing selective-judging methods can overstate what is safe: a naive certificate can claim 98% automation while its realized error exceeds the budget in 17.5% of task resamples. We introduce a task-level bootstrap certificate that is valid in every regime we test while matching the naive certificate's coverage; finite-sample cluster-valid alternatives certify nothing at realistic task counts. Under this certificate, a 4B logprob judge trained with SFT and reject-weighted GRPO certifies 0.30-0.59 of evaluation on tool-use and web corpora at alpha=0.1, the only judge, among strongly elicited frontier models, certifying on both headline corpora. Certified coverage is predictable before training from base rate and discrimination alone (leave-one-corpus-out R^2=0.96). Finally, the certificate doubles as a self-training filter: pseudo-labels harvested inside certified regions have contamination bounded by alpha by construction (realized 0.000-0.041 across six harvests), letting a judge enter an unseen domain at in-domain strength with zero target training labels.

Chengguang Gan, Yunhao Liang, Qinghao Zhang +1

Independent Researcher · University of Chinese Academy of Sciences · Pusan National University +1

LLM-as-a-Judge Selective Prediction LLM Agent Evaluation

Sep 28, 2026cs.CV

PlaylistEval: Can Video-Language Judges Be Trusted at Day Scale and Beyond?

Video-language models are increasingly used as judges of video understanding, both for evaluating model outputs and for training reward models. Whether their judgments remain reliable when the evidence is buried in day-long videos has yet to be established. Existing benchmarks cannot answer this. Their videos are typically only a few minutes long, many answer pairs can be separated from the transcript alone, and collecting human judgments does not scale to ultra-long videos. We introduce PlaylistEval, an agentic framework that builds video-language judge benchmarks over 100-hour playlist collection without human annotation. It automatically generates questions with paired answers whose differences are controlled by causal degradation, so that every pair demands retrieval across the collection. The resulting benchmark contains 630 pairs across seven domains spanning both static and dynamic knowledge, and on a stratified subset of 152 pairs it agrees with human judgments 93.0% of the time (IAA 0.781). Evaluating 17 omnimodal and multimodal models from eight families reveals that frontier judges reach only 75.4% pairwise accuracy, while open-source judge models perform far behind. We further show that both retrieval and final judgment depend on using multiple modalities, and that judge accuracy degrades as the playlist set grows. We release our pipeline, benchmark, and evaluation code at https://playlisteval.github.io.

Shayekh Bin Islam, Hwanjun Song

KAIST

LLM-as-a-Judge Video-Language Model Evaluation Cross-Modal Retrieval Long-Video Understanding Multimodal Model Evaluation

Sep 28, 2026cs.LG

Frozen Judges, Moving Agents: Version-Dependent LLM-Judge Error and the Limits of Judge-Assisted Agent Evaluation

Language-model judges compare agent upgrades with their predecessors, but a fixed judge can make version-dependent mistakes. We analyze 35 public coding-agent submissions (20 prespecified version pairs on 250 SWE-bench Verified issues), two customer-service agents (155 tau-bench tasks), and 1,106 expert-labeled AgentRewardBench trajectories. An upstream outage left three judges for the primary SWE-bench analysis (8,743 aligned cells); the fourth is descriptive. All three coding-agent judges and all four tau-bench judges reject task-conditioned error invariance after multiplicity adjustment. On SWE-bench, 32 of 60 judge-by-pair units have a detectable differential comparison component; eight judge-only intervals declare upgrades that execution-based intervals cannot establish, despite rank correlations of 0.71-0.79. In tau-bench, one judge reverses a nine-point reference-reward gap by penalizing a procedural habit the reward ignores. False acceptance of failed coding patches rises with agent capability conditional on task and execution outcome, while a task-solvability prediction reverses sign across domains. A separately fixed post-submission OpenHands follow-up on the same 250 issues (eight configurations, 1,981 three-judge cells) reproduces the capability/false-acceptance association (mean Spearman +0.944, exact p=0.000099) and decreasing Youden contrast (mean -0.937, p=0.000397); this is observational, not a new-task replication. Transporting old-version calibration raises SWE-bench comparison error from 3.8 to 19.5 points, with 24.6% undefined bootstrap ratios. A paired audit saves only 5% in interval width at 80 labeled tasks. A randomized self-report test is negative (three adjusted p-values=1.0). These results favor paired audits of current outputs over judge-only release decisions or transported calibration; independent human patch review remains pending.

Jiapeng Li

LLM-as-a-Judge AI Agent Auditing LLM Agent Evaluation Benchmark Validity

Sep 27, 2026cs.AI

Designing Reliable LLM-as-a-Judge Measurement Systems for Multi-Turn Business Agents

Many LLM-as-a-judge evaluations score fixed outputs under a fixed task definition. Production multi-turn business agents instead require a maintained measurement system: correctness depends on business-specific facts and procedures, outcomes emerge across turns, and failures must be attributed to either agent capability or missing business knowledge before they are actionable. We present an integrated methodology spanning evaluation specification, modular LLM judges, intent-preserving user simulation, and human-in-the-loop governance. The specification defines conversation-level end states and actionable failure ownership. Atomic judges share versioned evidence and feed an explicit aggregation graph. The simulator is released only after task-preservation and stability checks. Independent human audits estimate measurement fidelity, renew tiered reference sets, and route disagreements to label correction, guideline revision, or judge improvement. Production studies show that system-level fidelity improved across repeated audits, that human reviewers and automated judges improved together under the shared feedback loop, and that their combined workflow had the strongest descriptive performance in both reported task-completion settings. Because the studies are observational and the human reference itself required revision, these findings demonstrate operational usefulness rather than causal or universal superiority. The contribution is a practical framework for making multi-turn agent measurement reliable, actionable, and maintainable as the evaluated system and its evidence evolve.

Kaiwen Luo, Ming Gao

Meta

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

Sep 27, 2026cs.AI

Curating Merchant-Matching Training Data with Two Confidence-Gated Local LLM Judges

Merchant matching resolves a noisy payment descriptor to a retrieved merchant entity or returns no match. A key challenge in curating training labels is distinguishing teacher abstention from evidence that no acceptable entity exists: false no-match labels contaminate pseudo-labeled data, while conservative labeling reduces coverage. We investigate whether agreement between two local large language model judges improves pseudo-label reliability. A label is retained only when the judges agree, with separate ordered thresholds for selections and abstentions that guarantee disjoint positive and negative label sets. Retrospective replay on 2,000 expert-annotated queries shows that higher selection thresholds can improve positive-label purity, whereas higher abstention thresholds increase false no-match labels. At thresholds (0.86, 0.80), Muse Glimmer 30B and Gemma 4 31B jointly label 1,633 queries (81.7% coverage) at 96.88% purity; positive and negative purities are 99.47% and 93.38%. This exceeds either constituent model at the same thresholds by more than two percentage points, with lower coverage. A split-half check finds only 0.14 percentage points of threshold-selection optimism. A symmetric threshold of 0.86 adds 40 erroneous no-match labels, while 46 false abstentions persist even with no confidence threshold. Across five matched within-model comparisons, higher reasoning effort yields no clear F0.5 gain and increases median latency by 1.8-5.0 times. These results motivate separate thresholding and auditing for positive and negative pseudo-labels. The study establishes label purity, not student utility; fresh-data curation and student fine-tuning remain necessary to demonstrate downstream value.

Donghao Huang, Jinling Pei, Zhaoxia Wang

Research and Development, Mastercard, Arlington, VA, USA · School of Computing and Information Systems, Singapore Management University, Singapore

LLM-as-a-Judge Selective Prediction Pseudo-Labeling

Sep 27, 2026cs.AI

Learning Strategies to Break Judges

As AI agents surpass human performance, it becomes exceedingly hard for system designers to evaluate them directly and understand their failure modes. Consequently, agents themselves are being deployed extensively to evaluate, judge, and provide feedback on model traces. But this raises an important question: how can we trust the judge? In this work, we propose an agent-guided method to find weaknesses of agentic judges that expose interpretable failure mechanisms. Our method focuses on mathematical reasoning and proceeds in two stages: first, we deploy adversarial agents to mutate a set of sound proofs by introducing errors, attempting to misguide judges---in other words, injecting errors that judges are unable to catch. Then, we distill these attempts into a small set of mutation strategies which allow us to analyze the failure modes of the judges. To ensure that these strategies are not overfit to the initial set of proofs, we evaluate them by applying the mutation strategies to a held-out set of proofs and querying the same judge. We deploy our method on GPT-5.6-sol and Claude Opus 5, paired with their agent orchestrators, Codex and Claude Code, respectively. These are used both as mutators to introduce errors and as judges to evaluate correctness of mathematical reasoning. We find that across all the agentic judges, we are able to distill mutation strategies that consistently bypass their evaluations, thereby enabling us to ascertain actionable failure modes. Our analysis also reveals that judge reliability degrades at the frontier: errors in Olympiad-level proofs or graduate-level mathematical texts are detected more consistently, whereas flaws in research-level manuscripts are more likely to escape detection.

Guruprerana Shabadi, Aaditya Naik, Rajeev Alur +1

University of Pennsylvania · Oracle

LLM-as-a-Judge AI Agent Auditing

Sep 27, 2026cs.CL

Pinned and Still Unstable: Within-Judge Verdict Variance and the Noise Floor of LLM-as-Judge Leaderboards

Modern LLM evaluation assumes that pinning a judge to a fixed model version and decoding at temperature zero yields reproducible verdicts. We show this assumption fails as a property of how LLM-as-Judge is operationalized on cloud serving infrastructure, not of any particular model family. Across four frontier judges served via a single major enterprise cloud platform and three standard benchmarks (Arena-Hard, AlpacaEval 2, MT-Bench), identical inputs to the same temperature-zero judge, at a constant serving-reported model version, produce different verdicts across re-runs: per-item flip rates of roughly 5% on average and about 40% on the close-call items that decide leaderboard margins, with a per-judge magnitude spanning a 40x range (0.13% to nearly 10%). We introduce metrics tailored to this instability: per-item flip rate, a two-part stability profile (waver fraction and conditional intensity), and adjacency separability. For the principal judge the aggregate ranking is stable (0% top-K instability, 0% pooled winner flip); what degrades is precision: under a paired hierarchical bootstrap, roughly one-fifth to three-quarters of adjacent leaderboard positions are statistically indistinguishable, a noise floor driven mainly by finite prompt sampling rather than the judge. Across judges, leaderboards agree on the coarse ordering but diverge in the middle (Kendall's tau of 0.42-0.64 between Gemini and Sonnet judges on Arena-Hard, values sensitive to answers truncated at the generation cap). Of 15 expected head-to-head orderings we re-judge, 12 survive every re-run of the principal judge but only 8 survive every judge. Leaderboards thus report unhedged point estimates that overstate their precision. We propose a minimal, low-cost reporting protocol: several judge re-runs, published stability profiles and adjacency intervals, and results under at least two judges from different families.

Krishna Chytanya Ayyagari

Google

LLM Evaluation LLM-as-a-Judge LLM Reliability

Sep 25, 2026cs.AI

LLM Judge Validation Under Sparse Overlap: From Inference to Design

Validating an LLM-as-a-judge requires estimating its agreement with humans, yet annotation budgets rarely allow every item to be multiply labeled. We prove that this overlap sparsity is the first-order determinant of wrong deployment decisions: at 5% pairwise overlap, wrong-decision rates reach 25% and the probability of selecting the wrong best judge among ten candidates is 65%. The two actionable levers are overlap quantity and allocation. For quantity, we derive a minimum-overlap formula showing

ρ\geq 0.25

suffices for non-borderline judges while borderline cases remain fundamentally hard. For allocation, a zero-cost stratified scheme halves false-rejection rates relative to random sampling when strata are informative. We validate on 10 LLM judges across four evaluation matrices spanning visual assessment, causal reasoning, and summarization.

Junxuan Li, Arko Mukherjee, Soumyabrata Pal

Adobe · Adobe Research

Inter-Rater Reliability LLM Evaluation LLM-as-a-Judge

Sep 24, 2026cs.CL

JEV vs. LLMs as Rubric Judges: Cheaper, Faster, and Wrong in the Same Places

We ask whether Jev, a typed classifier that returns probabilities over permitted answers without generating text, can replace an LLM rubric judge. We compare it with three flash-tier LLM judges on nine panels drawn from seven benchmarks, giving every judge identical criterion texts. Jev's accuracy differs significantly from an LLM judge's in only 8 of 27 paired comparisons, ahead mostly on binary criteria and behind only on graded ones, and most of the other comparisons are inconclusive. Summed over the nine panels, the LLM judges, called once per criterion, cost 29 to 325 times as much as Jev and took 30 to 220 times as long. On graded criteria all four judges agree more with one another than with the labels and mostly assign lower levels than the raters. One of several observational accounts is that raters followed scale conventions our criterion texts omit. Jev's confidence ranks its own errors on most panels, which should make a cheap classifier the ideal first stage of a cascade that defers its uncertain verdicts to an LLM judge. Correlated errors undo that advantage. The LLM judges repeat nearly all of Jev's most confident errors, so a cascade replayed on the recorded verdicts lowers cost but gains at most 1.5 points over the best single judge with cross-fitted thresholds, and at most 2.0 even with oracle thresholds.

Delip Rao, Chris Callison-Burch

University of Pennsylvania

LLM Inference Efficiency LLM Evaluation LLM-as-a-Judge Selective Prediction Classification Rubric-Based Evaluation

Sep 24, 2026cs.CL

Where LLM Graders Succeed and Break: Evidence from Two Computer-Science Exams

One long-form exam in a large course costs hundreds of grader-hours, and qualified graders are scarce; LLM graders are a tempting alternative. To show its pitfalls we grade a practical Computer Vision exam (

570

dual-graded students) under

171

configurations spanning closed and open-weights models; the best reaches mean absolute error

1.64/35

, below the

2.61/35

two human graders achieve against each other. The catch is the prompt: a short ''strict grader'' preamble drives

14

of

17

open-weights models out of the graded band (

\text{MAE} \ge 8

), three stopping grading altogether. The damage traces to the preamble's two credit-withholding sentences, not to tone or model scale; one of them, ''never give partial credit'', alone makes two of three probed models stop grading. The closed flagships of three vendors shift calibration under it but stay in the band. In

162

further configurations on a second, independent Machine Learning exam from another course (

1{,}038

dual-graded students), the preamble worsens ten models, moving three out of the band into collapse and one into refusal, yet improves seven whose neutral prompts over-mark: the vulnerability replicates, but its direction is exam-specific. Light LoRA fine-tuning repairs it: one adapter on the two exams' pooled

\sim 3{,}900

graded examples brings five small open models to parity or better with a human grader in agreement with the grader pair, and sensitivity to the three harsh personas nearly vanishes (

\le 0.32

MAE). We release the anonymised dataset, full ablation grid, and grading, fine-tuning and analysis pipelines.

Ali Habibullah, Yazan Alshoibi, Mohammad Alshiekh +2

KAUST Academy Computer, Electrical & Mathematical Sciences & Engineering (CEMSE) King Abdullah University of Science and Technology (KAUST) Thuwal, Saudi Arabia · Visual Artificial Intelligence Laboratory Oxford Brookes University · KAUST Academy & CEMSE, KAUST +1

Prompt Sensitivity LLM-as-a-Judge LLM Auditing Language Model Calibration Automated Grading

Sep 23, 2026cs.CR

Don't Read the Log: Execution Traces Contaminate Verifiers in Video-Generation Agents

Agentic video-generation systems close a loop between a generator and a verifier: an LLM plans shots, calls a text-to-video model, and a multimodal judge decides whether the result satisfies the request. To diagnose where a long workflow fails, recent harnesses deliberately show the judge more than the video-the agent's execution trace, its plan, the narration it synthesized. We ask whether this auxiliary text moves the judge's verdict on purely \emph{visual} requirements, holding the frames fixed. On a benchmark of 109 generated two-event clips with manual labels, in which the requested event is either visibly completed or visibly missing, a trace that reports a successful tool call makes three open-weight Qwen-VL judges (7B, 8B, 32B) accept

78

--

90\%

of the failures, up from

7

--

19\%

without text, and a contradicting trace makes them reject up to

100\%

of correct clips; an instruction to ``use only the frames'' does not remove the effect. Frontier closed judges are essentially unmoved on the same clips, showing that the vulnerability is a property of the judge's learned trust in tool logs rather than of the task. Plan-derived text carries no clip-specific information, so it can only shift a judge's operating point, and in a repair loop that shift becomes a cap on the true pass rate that no repair policy can exceed; the cap matches simulation to two decimals. In the loop, contamination is exploited without any adversarial agent: an honest LLM planner that always regenerates ends with a judge pass rate of

1.00

and a human-labelled pass rate of

0.28

, and a pipeline in which a cheap checker writes its verdict into the trace launders that checker's errors into a stronger final judge (

0.69

false accepts).

Jian Xu

RIKEN

LLM-as-a-Judge AI Agent Auditing Agentic Video Generation Multimodal Large Language Models LLM Agent Reliability Video Generation Evaluation

Sep 23, 2026cs.CL

Can Jev Judge Radiology Reports? Evaluating a System One Model for Clinical Factuality

An AI-generated radiology report can resemble a physician's report while omitting an abnormality, adding an unsupported finding, or reversing its presence. Measuring these factual differences is essential for evaluating report generators. We study Jev, a System One decision model, as a simple, low-cost judge of agreement with physician-written reference reports. Our evaluator checks whether each statement is supported by the other report and combines these judgments in both directions to capture unsupported claims and omissions. A single-question configuration reaches Kendall correlations of 0.573 on RadEvalX and 0.398 on RadEvalExpert with expert error counts, outperforming an open natural language inference judge under matched decomposition and aggregation. One support question per statement retains similar expert agreement to seven while using 43-45% fewer judgment input tokens. At the documented API price, judgments cost under three cents per hundred report pairs, excluding local decomposition. In a separate controlled-error test, Jev detects false negation with an AUROC of 0.977. Local RadMatch achieves stronger agreement on clinically significant errors in both expert datasets and on total errors in the shared RadEvalExpert subset. Finding-count and error-scope analyses show that benchmark agreement reflects report size and error definitions as well as medical error detection. These results support Jev as a practical judgment component for measuring factual differences in generated radiology reports and identify where more elaborate evaluation remains valuable.

Jiaju Huang, Hao Yang, Xinyu Ma +6

Intelligent Medical Computing Laboratory, Faculty of Applied Sciences, Macao Polytechnic University · Department of Radiology, Netherlands Cancer Institute, Amsterdam, the Netherlands · Department of Radiology and Nuclear Medicine, Radboud University Medical Center, Nijmegen, the Netherlands

Radiology Report Generation LLM-as-a-Judge Factual Consistency Evaluation Radiology Report Evaluation

Sep 22, 2026cs.AI

JEV-as-a-Judge: Accept When Confident, Escalate When Unsure

LLM-as-a-judge scales evaluation, but reasoning judges are slow and costly. We study JEV-as-a-Judge: evaluation with JEV, a decision-only judge that returns label probabilities instead of text, and whose confidence decides whether to accept its verdict or escalate to a reasoning judge. Against sixteen generative and reward-model judges, with blinded human adjudication, JEV comes within three points of GPT-6 wherever a verdict can be read off the text, at 0.36% of its fee and a 0.15-second median latency, and falls behind where the verdict must be derived, as in math, code, and logic. Its confidence marks this boundary. With a threshold frozen in advance, accepting confident verdicts and escalating the rest is 0.9 points more accurate than GPT-6 on 1,610 held-out pairs at 41% of its fee, and in a pre-specified live test on two new workloads the cascade matches GPT-6's accuracy exactly. Confidence routing weakens on style-adversarial pairs and reference-free prose; we close with a simple recipe for validating thresholds locally.

Yubo Li, Yidi Miao, Ramayya Krishnan +1

Carnegie Mellon University

LLM-as-a-Judge Language Model Generation Evaluation Selective Prediction LLM Uncertainty Estimation Automated Evaluation

Sep 21, 2026cs.IR

GroundedGEO: Auditing the Evidence Gap in Generative Search Rankings

Generative search systems rank products and services for consequential decisions, and publishers can cheaply make candidate text look relevant. Yet evidence status is not a text property but a claim-evidence relation: text-only rankers and defenses cannot separate honest detailed content from fabricated detail, creating an identifiability gap. We audit this gap with an evidence-paired benchmark (50 e-commerce queries, 1,950 cases) and a claim-level reranker, GroundedGEO, that penalizes query-relevant claims lacking support in a supplied packet. Matched rich variants control format and volume; packet twins add attestations at fixed text, while thinned packets withdraw them. On the frozen listwise ranker Qwen2.5-7B, unsupported-rich variants show significant normalized rank gain over clean candidates (+0.065 to +0.092 across claim profiles, Holm-corrected), while supported and neutral controls do not; the effect is model-dependent (marginal on MiMo-v2.5, absent on GLM-5.3-Flash). On a frozen pointwise scorer, oracle evidence labels cut the unsupported-rich top-3 rate from 0.65 to 0.43 (laundering from 0.61 to 0.39) at lambda=40 with zero false suppression; packet twins restore the original rates without changing text. Against a 370-claim human gold, all tested automatic judges fail the preregistered reliability gate, although the best local judge retains 79-100% of oracle suppression with zero measured false suppression on protected arms. Separately, stripping attestation coverage increases false suppression by 0.307. These diagnostic effects identify two limits on the evidence channel: label quality and packet coverage. They do not validate an automatic defense, and interpretation of the adverse human-gold arm remains pending adjudication.

Yihan Xia, Huiling Fan, Kangrong Zhong +1

Shenzhen University Shenzhen, China

LLM-as-a-Judge LLM Reranking

Sep 21, 2026cs.CL

LLJ Cards: Best practices for the Use of LLMs as Judges

In recent years, large language models (LLMs) have emerged as a popular alternative for evaluation. Often referred to as LLMs as judges (LLJs), these systems have been widely adopted by researchers and practitioners across a broad range of measurement tasks, driven by their strong performance, scalability, and cost-effectiveness relative to human judgment. However, a growing body of work has shown that the use of LLJs raise concerns about their validity and reliability as evaluators. Existing efforts to address these challenges have largely focused on developing bias-mitigation techniques and refining prompting strategies. While these approaches represent an important step forward, they primarily offer technical fixes and leave a more fundamental challenge unaddressed: the lack of standardized, transparent, and reproducible evaluation practices. In this paper, we introduce LLJ Cards, a framework that synthesizes best practices from measurement theory, natural language generation, and machine learning literature into practical guidelines for LLJ-based evaluations. While LLJs offer a promising path toward scalable evaluation, their effective use requires grounding in rigorous evaluation principles to ensure validity, reliability, and reproducibility. LLJ Cards addresses this need by providing a structured framework for applying these principles in the design and reporting of automated evaluations.

Khaoula Chehbouni, Melina Medjdoub, Florian Carichon +2

McGill University · Mila - Quebec AI Institute · Concordia University

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

Sep 21, 2026cs.CV

ChartJudgeBench: Evaluating LMM Judges for Chart-to-Code Generation

Building strong chart-to-code systems increasingly relies on reinforcement learning, whose effectiveness depends critically on the quality of the reward signal. Large Multimodal Models (LMMs) play a natural critical role in jointly assessing chart visual appearance and task requirements. They are therefore increasingly used as visual critics and reward models, yet their reliability as judges remains largely unexplored. To this end, we introduce ChartJudgeBench, a diagnostic vision-language benchmark for assessing LMM judges in chart-to-code workflows. It includes 1,003 Chart Perception Alignment (CPA) instances for pairwise chart comparison and 650 Chart Reasoning Judgment (CRJ) instances for binary Accept/Reject verification in Chart Reproduction and Chart Editing. Together, these tasks emulate the core judging decisions required in agentic refinement and RL-based chart optimization. Our evaluation of strong LMMs reveals four systematic limitations: (i) positional bias in pairwise comparison, (ii) a strong tendency to overpredict Accept, (iii) difficulty in matching visual styles and aesthetics, and (iv) an unexpected leniency bias in RL-trained models. These findings show that current LMM judges require explicit reliability validation before being used as critics or reward models in chart-to-code optimization. The code and data are available on ChartJudgeBench.

Lijian Wu, Henry Hengyuan Zhao, Zijian Zhang +3

CSU-JPG, Central South University · National University of Singapore

VLM Evaluation LLM Evaluation LLM-as-a-Judge Code Generation LLM Reliability Vision-Language Model-as-a-Judge

Sep 21, 2026stat.ML

OSCAR: Order-aware Scoring and Calibration for AI Rankings

Judge-specific sensitivity is useful for aggregating pairwise LLM evaluations, but its interpretation depends on which systematic presentation effects the ranking model includes. We introduce OSCAR, an order-aware framework for scoring and calibrating AI rankings, and study position as one such effect. In released judgments from 18 evaluators, the all-response A-minus-B score difference ranges from

-63.11

to

98.31

percentage points. Matching question text, response texts, candidate identities, and judge within the released table gives an overall difference of

24.22

points (95% interval

[22.90,25.54]

), conditional on the released text mapping. A controlled calculation isolates the potential consequence: with true sensitivity fixed at one, omitting a position intercept of four reduces the population-optimal slope to

0.0771

. We extend sensitivity-based ranking with judge-specific position, length, and family terms, characterize local omission-induced displacement and an identification failure, and propagate prompt-cluster uncertainty to adjusted comparisons. Across four released datasets, position provides the largest stand-alone predictive improvement. Refitting bootstrap comparisons show more selective gains from the full model over position-only adjustment. In dependent binary simulations, adjusting both the mean and covariance yields 94.4--95.2% coverage; correcting either alone is insufficient. At

N=10{,}000

, OSCAR reduces mean neutral-target RMSE from

0.1158

under the sensitivity-only model to

0.0237

.

You Liu, Yue Liu, Quanchao Lu +1

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

Sep 18, 2026cs.CL

How Many Humans Are 32 LLM Judges Worth?

A panel's human-equivalent size is target-specific. Matching a fixed 32-judge panel to empirical human label distributions on three ChaosNLI tasks yields two distinct effective sizes: distributional-error matching gives

ν_{\mathrm{MSE}}=2.304

,

3.750

, and

3.445

, whereas spectral matching gives

ν_H=4.242

,

6.459

, and

6.499

, a gap of

1.72

--

1.89\times

; a binary-error diagnostic credits the same panels with only

1.971

--

2.227

effective votes. Extrapolating the distributional-error curve at fixed squared mean residual, mean member variance, and normalized mean covariance gives asymptotes of

2.392

,

3.990

, and

3.655

, with 32 judges already reaching

94.0

--

96.3\%

. An exact spectral identity explains the gap: error depends on member energy and on the orientation of residual variation relative to averaging, information that the participation ratio (PR) discards. A realizable hard-label construction confirms that higher spectral diversity can coexist with worse distribution recovery even under equal member energies and nonnegative correlations, and the consensus direction retains

γ_{\mathrm{co}}=43.8\%

,

33.7\%

, and

35.9\%

of centered residual variance. An external check on CC-1000, a 1,000-item Civil Comments subset with a different panel, gives

ν_H=2.84

. For panel choice, we establish an existence result and one feasible path: exhaustive enumeration at

k\in\{5,7\}

shows that panels beating the accuracy-top-

k

baseline on both accuracy and

ν_H

always exist, and greedily swapping at most two members reaches

24.8

--

56.0\%

higher

ν_H

at

0.10

--

1.10

percentage points higher accuracy. Our dataset and code are available at https://github.com/Chao1208/32judges-votes.

Chao Li, Yingying Yu, Yunfeng Li

Tsinghua University · University College London · Hebei University of Economics and Business

Annotator Disagreement LLM-as-a-Judge Language Model Ensembles Ensemble Diversity

Sep 17, 2026cs.CL

VākQA: A Benchmark and Evaluation Study for Telugu Spoken Factoid Question Answering

Question answering has advanced rapidly with large language models, but predominantly for high-resource languages, in both text and spoken settings. Spoken question answering (SQA) benchmark for Telugu remains unexplored, and the reliability of automatic evaluation in this setting remains unquantified. We introduce VākQA, a Telugu SQA benchmark of 2,001 factoid question-answer pairs across six domains, with 2.53 hours of speech audio, bilingual transcriptions, and human-verified reference answers. We first validate evaluation methods against human judgements: Gemini-as-a-judge best approximates human ratings but is non-uniformly strict, while open-weight judges systematically penalize correct Telugu answers that differ in surface form from the reference. Using this validated setup, we benchmark proprietary and open-weight models across input modality, language, and domain. We observe that Telugu phrasing retains cultural specificity that is lost in translation, speech input introduces phonetic confusions that alter question meaning, and cascaded ASR-MT errors compound progressively. VākQA is publicly released.

Bhavana Akkiraju, Ravi Sastry Kolluru, Sri Charan D +3

International Institute of Information Technology Hyderabad, India

Audio QA LLM-as-a-Judge Question Answering Multilingual QA Spoken QA

Sep 16, 2026cs.CL

I code or AI code: A comparative evaluation of AI-rated scores in classroom observations

Classroom observations are widely recognized as a key tool for establishing benchmarks of education quality and guiding pedagogical improvement, yet they remain resource-intensive and dependent on trained observers. This study evaluated the feasibility of using a LLM (GPT-5 model) to score teacher-child interactions in early childhood classrooms, benchmarked against human raters. The study analyzed 87 video-recorded observations from 38 classrooms across 30 kindergartens in Hong Kong. Using observation transcripts, the AI model was configured to apply the full Classroom Assessment Scoring System (CLASS) framework. AI-rated scores were then compared with human ratings by examining correlations and differences in mean scores of the CLASS domains and dimensions. The results showed greater convergence between AI and raters for the Emotional Support domain and, in particular, the Quality of Feedback dimension, which captures how teachers use feedback to extend children's learning. Greater divergence emerged for interactions that were more procedural or context-dependent, particularly within the Classroom Organization and Instructional Support domains. These findings suggest that transcript-based AI scoring may capture some of the relative variation in teacher-child interactions but cannot yet reproduce calibrated human judgements consistently across the full CLASS framework. AI-assisted observation may therefore be more appropriate as a preliminary screening tool rather than as a replacement for trained observers, providing teachers with evidence for reflection rather than high-stakes evaluation. Future research should examine whether domain-specific training and incorporation of contextual and visual information can improve alignment between AI and human rated scores.

Y. Fong, J. Xiang, T. Y. D. Chan +2

Department of Early Childhood Education, The Education University of Hong Kong, 10 Lo Ping Road, Tai Po, Hong Kong SAR, China · Yew Chung College of Early Childhood Education, 2 Tin Wan Hill Road, Tin Wan, Hong Kong SAR, China

LLM-as-a-Judge Educational Assessment AI in Education Automated Evaluation Generative AI Evaluation

Sep 16, 2026cs.CL

Beyond Accuracy: How Procedural Traces Shift the Decision Criterion of LLM Overseers

Organizations increasingly use oversight loops where one large language model (LLM) audits another's outputs alongside procedural traces of claimed steps. A common concern about such LLM-as-a-judge pipelines is that detailed traces make overseers gullible. Using signal detection theory, we audit five LLM overseers on 19 compliance tasks (4,551 analyzed judgments), varying only trace detail and evidence labeling. With disconfirming evidence always visible, error detection remains near ceiling. Instead, elaborate traces shift the decision criterion toward rejection, increasing false alarms in susceptible overseers. Without option labels, human-validated reason coding shows about 60% of false alarms cite an inability to tie evidence to its option. Labels eliminate this stated reason, yet residual rejection of correct work persists in those overseers and rises with trace detail. Procedural traces thus act as governance artifacts that shape oversight decisions. AI auditors should be evaluated by their decision criterion and false-alarm behavior, alongside accuracy.

Zihan Chen, Di Zhu, Lei Zheng +1

Stevens Institute of Technology Hoboken, NJ, USA · University of Massachusetts Boston Boston, MA, USA · Stony Brook University Stony Brook, NY, USA

LLM-as-a-Judge LLM Auditing

Sep 16, 2026cs.CL

A Calibrated Instrument for Measuring How Inference Optimizations Affect Output Quality

Large language model optimization is an active research area, spanning quantization of model weights, early-exit methods for skipping layers, and speculative decoding. Each track uses its own quality measures, typically an idiosyncratic benchmark score. Few approach the measurement precision required by other scientific disciplines. We propose a rigorous methodology for measuring output quality, suitable for cross-system and cross-technique comparison. We score outputs with an LLM as a judge, but calibrate the judge formally: we compare its scores on two ordinary runs of a model given the same prompts, verifying that it shows no systematic preference between statistically equivalent outputs and measuring its per-sample noise. Each design also includes a 'null' condition, provably identical in distribution to the unmodified model, whose measured difference must be zero. With this one instrument we measure several acceleration techniques on the same prompts, so their quality costs can be compared. Perceived quality proves highly dependent on the domain of discourse. A 4-bit model was indistinguishable from its 16-bit original down to our design's +/-0.3-point resolution, in English prose and Chinese alike. At 3-bit precision the same prompts lost 0.5 points in English prose, 0.9 in Chinese, and 1.1 on multi-step math; early exit that cost 0.7 points on prose cost 2.5 on math, cutting correctly solved problems from 19 of 27 to 6. The pattern held for models from Alibaba and from Meta, but not its magnitude: the same quantizer cost Meta's model 1.8 points where it cost Alibaba's 0.7. A model's certainty about a token predicts how likely it is to differ from the full model's choice, but not how much that difference affects judged quality, so acceptance rules relying on certainty cannot distinguish errors that matter from errors that don't.

Jerry Kaplan

Adjunct Lecturer, Computer Science and Masters in International Policy, Stanford University

LLM Quantization LLM Evaluation LLM-as-a-Judge Language Model Generation Evaluation LLM Inference Acceleration

Sep 15, 2026cs.CL

Who Judges Matters: Measuring Family-Conditioned Preference in LLM-as-Judge Panels

Who the judge is can affect an LLM-as-judge result, but measuring that effect without confusing it with candidate quality is difficult. We study four open-weight families (Llama 3.1, Qwen 2.5, Gemma 2, and Yi 1.5) in a fully crossed pairwise design with 9,312 judgments. A common per-family statistic is strongly confounded with candidate quality and correlates with Bradley-Terry ability at r = 0.95. We derive a corrected estimator that holds the candidate family fixed and compares judges. All four families then show a positive same-family lift (3.4-8.4 percentage points), with global FPS 0.067 (95% CI [0.053, 0.084], permutation p = 0.0002). The effect remains under panel-based quality controls, an independent human-consensus anchor, and a float16 judging replication. Judge-side likelihood is closely related to the effect: adding likelihood advantage reduces the controlled coefficient by 61%, which we treat as descriptive attenuation rather than causal mediation. Position is a separate failure mode. Across the panel, 55.4% of AB/BA pairs reverse, and reversal above 50% is incompatible with a simple independent content-noise model. Relative to a family-balanced reference, panel composition changes 18.5% of pairwise outcomes. A complete reproducibility archive has been prepared for public release.

David Ababio Awuni, Luke E. K. Achenie, Benjamin Tei Partey +2

LLM-as-a-Judge Pairwise Preference Evaluation Language Model Bias Evaluation Bias in LLM-as-a-Judge

Sep 15, 2026cs.CL

Beyond the Name: Demographic Leakage in De-Identified Résumés and Evaluation Artifacts in LLM Bias Audits

De-identified résumé screening assumes that redacting explicit fields prevents ethnocultural inference; however, recent audits attribute residual leakage to declared languages. We investigate whether eliminating language fields resolves this leakage across nine open-weight models and 620 counterfactual résumés. By holding language attributes strictly identical, we isolate unstructured prose across five ethnocultural conditions and three cue-salience tiers. Target-group recovery averages 0.757 overall and saturates at 1.000 under high salience, demonstrating that non-language prose sustains demographic inference. Crucially, models diverge only under faint cues (0.086-0.690), establishing salience as an essential evaluation axis. Furthermore, pairwise LLM-as-a-judge outcomes are highly sensitive to evaluation design: forbidding ties yields an apparent selection-rate ratio of 0.39 alongside strong position and content effects, whereas permitting ties produces near-universal ties for most models (

\ge94\%

). Downstream scoring shows only very small between-condition differences, highlighting the need to distinguish demographic signals recoverable from résumé content from effects introduced by the evaluation protocol.

Qiangju Chen, Yang Xiao

Macquarie University · The University of Melbourne

LLM-as-a-Judge Language Model Bias Evaluation Demographic Bias in Language Models

Sep 15, 2026cs.AI

Skill-based Agentic Evaluation for Real-time Data Science Tasks

We present a framework for evaluating data-science agents on live, continuously updated data using executable ground truth and format-agnostic factoid scoring. Consider this example query: "what were last week's audience sizes"---the reference answer changes as the underlying data changes, so static references become outdated and standard LLM-as-a-judge pipelines cannot verify responses against a fixed ground truth. Our central contribution, ground-truth-as-code, encodes each expected answer as an executable reference function that recomputes the answer directly from live data at evaluation time, ensuring the reference remains consistent with the system it describes. We combine this with a factoid-level, format-agnostic judge that decomposes both the agent's response and the computed ground truth into atomic claims and scores precision, recall, and accuracy over them, irrespective of the response format (prose, list, table, HTML, etc.). The approach is applicable to agents whose expected outputs can be expressed as executable data computations. We validate the framework through a human--LLM agreement study on an internally developed machine learning skill deployed in production, using a synthetic database constructed to reproduce production schemas and entity relationships. Relative to a natural-language ground-truth baseline, our method achieves a 29% improvement in the Matthews Correlation Coefficient (MCC)---a class-balanced measure of agreement between expert annotators and LLM-as-a-judge predictions---and a 16% reduction in token consumption per test case, while a self-directed baseline lacking explicit ground truth is anti-correlated with human judgment. Agents that perform multi-source data integration and computation over non-stationary data are routinely deployed in industry; we propose ground-truth-as-code as a practical methodology for their evaluation.

Aniruddha Tamhane, Raghavendra Addanki, Ayushi Aggarwal +4

Adobe

LLM-as-a-Judge LLM Agent Evaluation Agent Evaluation Data Analysis Agents