LLM-as-a-Judge

LLM judges, models that score another system's output, can be gamed by the systems they score. Recent work identifies one defence that works: the judge solves the task itself first and commits to that answer, then accepts a candidate only if the two match. We call this commit-first judging, and ask whether shipped software implements it, and what it costs. We audit the default judge configurations of eight widely used evaluation frameworks. Of the 24 configurations in scope, none implement it. Nine implement a variant the literature measures as ineffective, and share one ancestor prompt, traceable through a copied typographical error. In a controlled experiment, an ordinary best-of-N search with no access to correct answers optimises code against one of these configurations, used exactly as documented. On an interval merging task the judge accepted 90 of 96 candidates in one seed and 93 of 96 in the other; every accepted candidate passed every test the search could see and failed a held-out suite it could not. The judge identified the defective line and cited it as grounds for a perfect score. Commit-first judging removed the effect: 0 of 96 in both seeds. On a second task it made matters worse in both seeds: the judge's committed answer was wrong, and in one seed the population converged on it. This is our main finding. Commit-first judging does not remove the anchor that gets gamed, it moves it from the candidate to the judge's own answer, so evaluation is only as good as the judge is at the task. That precondition is cheap to measure in advance, and is task local rather than scale dependent: a smaller judge solved a task the frontier judge failed and resisted gaming where it did not. We also validate our own instruments: five of fifteen claims in our criteria were wrong against verbatim sources, and two held-out checks were unjustified by their specifications.

Evaluations that use a large language model (LLM) as a judge have begun to borrow reliability statistics from classical test theory and its extensions. We examine three such statistics that need one administration and no gold labels. None of them can isolate the judge, because one judge under one prompt supplies no variance component of its own. Claude Haiku 4.5 judged 210 constructed short answers against ten-element checklists. On the 180 with parsed verdicts, the Kuder-Richardson coefficient (KR-20) came out at 0.5223 on the judge's verdicts and 0.5231 on error-free gold verdicts. In simulation, bank design alone moves KR-20 from 0.01 to 0.68 at the judge's measured 4.72% error rate. The dependability index $Φ(λ)$, a ratio of mean squared distances from the pass mark, sits 0.22 to 0.38 below the judge's accuracy against gold and returns 0.54 to 0.68 on error-free gold verdicts. Livingston-Lewis accuracy treats the rubric elements as a sample, and at a pass mark of five elements it credits error-free gold scores with 0.78, close to the judge's 0.81. A statement about the judge therefore needs gold labels or a varied scorer facet, and a reliability ratio needs the bank's spread beside it. One of the four closest judge-evaluation papers varies the prompt and still reads a reliability below 0.7 as a sign that a model cannot serve as a judge, although that reliability moves with the spread of the samples scored. We derive a decision table and four reporting lines from these two rules.

Multi-Agent Debate (MAD) has been widely adopted to improve LLM-based evaluation by prompting multiple agents to negotiate and reach a consensus. However, for subjective rubric-based scoring, inter-agent agreement does not guarantee alignment with human judgments. In this paper, we compare a single-judge baseline against a consensus-based MAD protocol on subjective evaluation tasks and design three ablations to isolate the impact of role prompting, multi-round interaction, and explicit score sharing. Evaluations across six LLMs show that the single-judge baseline achieves the strongest human alignment on average across six judge models, whereas MAD shows degradation in human alignment on both tasks. Our ablations demonstrate that this performance drop stems primarily from asymmetric role prompting rather than the interaction itself. Specifically, assigning a strict judge role introduces a systematic downward bias that the consensus process fails to correct. The central finding is that this bias reflects strict-stance dominance beyond averaging: the consensus score falls well beyond the arithmetic midpoint of the standalone strict and lenient conditions, rather than averaging them out. Removing role asymmetry (Symmetric MAD) largely recovers baseline performance, while masking peer scores widens inter-agent disagreement on average and worsens average human alignment. These findings demonstrate that multi-agent consensus can enforce artificial agreement at the expense of true human alignment, revealing a structural limitation in consensus-style, role-specialized MAD protocols for subjective scoring.

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

Large language models are increasingly used as scalable evaluators for open-ended tasks. However, many LLM judges derive query-specific criteria during scoring, leaving the evaluation requirements insufficiently specified and their coverage difficult to audit. Query-specific rubrics make these requirements explicit, but expert-written rubrics are costly to construct, while existing automatic methods typically rely on inference-time refinement or external supervision. We introduce GenRubric, a self-evolving framework that improves rubric generation from unlabeled queries without requiring additional human annotations during self-evolution. Our approach is based on rubric-induced self-consistency: independently sampled rubrics for the same query provide partial views of its latent evaluation requirements, and a comprehensive rubric should induce a response that generalizes across these complementary evaluation views. We implement this principle through reinforcement learning, combining a cross-rubric comprehensiveness signal with group-level and criterion-level rewards for rubric quality. We train GenRubric models at 4B, 8B, and 14B scales across multiple domains. Experiments on human-annotated rubric benchmarks show that self-evolution improves the agreement between evaluations induced by generated rubrics and those induced by expert-written rubrics. The improvements further generalize to held-out domains, demonstrating the potential of self-evolving rubric generation for scalable and query-specific LLM evaluation. Code and models are publicly available at https://github.com/foggpoy/GenRubric.

Evaluating research ideas generated by LLMs is difficult because their scientific value cannot be fully determined by objective criteria, and no single reference answer specifies what counts as a good idea. To address this challenge, we introduce Ideation Arena, a battle style platform that evaluates research ideas through pairwise human assessment. Ideation Arena evaluates ideas generated by 14 frontier LLMs and 5 research agent architectures built on 2 base models. To ensure a common starting point, Ideation Arena builds shared literature contexts from papers familiar to the participating researchers and provides the same contexts to all LLMs and agents. We collect over 6,000 double blind pairwise comparisons from 105 active computer science researchers and construct an Elo rating leaderboard of proposal-stage expert preferences in computer science under a shared closed-context protocol. We validate the rankings through interrater agreement and robustness analyses, showing that the leaderboard remains stable under changes in annotator composition and domain coverage. Our results show substantial variation in agent effectiveness, with some frameworks improving ideation quality over their backbones and others offering little benefit or even underperforming their base models. We further construct Ideation Arena Eval, a benchmark for assessing whether automated evaluators align with human preferences in research ideation. Experiments with current LLM judges show that they still cannot reliably reproduce expert preferences, with the best judge reaching 72.56% Soft Accuracy on Overall Quality. Our code, data, and leaderboards are available at https://github.com/foss12138/Research-Ideation-Arena.

LLM-judge audits assess bias by comparing ratings across matched conditions. Difference-in-differences designs compare two candidate responses within each item and then compare that contrast across a manipulated attribute. We show in closed form that this endpoint need not identify differential preference on the latent scale when ratings are bounded. A severity shift common to both responses produces an observed interaction whenever the scale bounds attenuate it unequally. We examine this problem in a pre-registered audit comprising 990 calls to a frozen pedagogy judge. The registered primary analysis did not detect an effect of the stated learner profile on scaffolding preference. The only nominally significant secondary interaction concerned productive struggle ($+0.378$ points; $p = 0.002$). A post-hoc construction with zero differential preference reproduced 79 to 85% of this interaction using the observed severity shift and the scale floor alone. The observed interaction therefore does not identify differential preference on the latent scale.

Self-evolving language models have recently emerged as a promising path toward superintelligence, with the advantage of reducing the cost of human supervision. While considerable progress has been made in verifiable domains, self-evolution in unverifiable domains remains less explored. We propose Judge co-adaptation from Zero data (J-Zero), a unified Challenger--Solver--Judge self-evolution framework that supports self-improvement across both domains. The Challenger and Solver co-evolve through an adversarial interaction: the Challenger generates increasingly difficult tasks, while the Solver learns to produce higher-quality responses to them. In parallel, the Judge co-adapts using preference pairs whose ordering is known in advance from how each response was produced, i.e., the Solver's answer over the Challenger's, and the Solver's decomposed-and-recombined answer over its one-shot answer, rather than from the Judge's own scores. J-Zero outperforms the baselines by an average of 4.2 points on verifiable and 8.0 points on unverifiable domains, and continues to improve through at least ten iterations, whereas the baselines degrade after two. Further analysis identifies Judge co-adaptation as the key driver of this sustained improvement.

Multi-agent systems (MAS) sometimes already have the potential to answer correctly, but still report a wrong answer. Explaining this outcome is difficult because generation, communication and final answer-selection rules usually change simultaneously. We conceptualize multi-agent reasoning as an evolutionary pipeline of candidate generation, peer communication and terminal selection, wherein consensus without quality control can exhibit patterns of memetic drift. We study two questions: (1) when an LLM judge provides effective selection pressure by supplying a signal of answer correctness for candidates generated in a multi-agent system, and (2) when using that signal improves the reported answer. To map judge reliability, we analysed 15,336 questions from MMLU-Pro, GPQA, MedXpertQA and MuSR, with Humanity's Last Exam analysed separately. To test these rules, we replayed 81,390 fixed candidate pools drawn from 16,278 questions across five benchmarks. We report three findings. (1) A correct answer is often already present among the generated candidates, but the system can still converge on and report a wrong answer. (2) Judge reliability is not a fixed trait of the model, but varies with the task, the generator and how rare the correct answer is. (3) Combining answer frequency with the judge's evaluation changed only the final answer-selection rule and raised accuracy from 63.82% to 70.82-70.95%, primarily by rescuing correct answers that were outnumbered by popular errors. In the systems studied here, the value of generating more candidates depends on whether those extra samples make correct answers present, frequent or recognisable. By isolating generation, recognition and selection, these findings establish a diagnostic basis for designing multi-agent architectures that protect generated correct answers from being lost.

Evaluating conversational voice agents at scale re- quires reliable assessment methods that capture both observ- able interaction quality and the contextual judgment typically provided by human evaluators. We investigate LLM-as-a-Judge evaluation by comparing human judgments with GPT-4.1 and GPT-5 on telecom and retail voice-agent conversations, across conversational quality and safety dimensions. The same interac- tions are scored under three evaluation configurations, p0, p1, and p2, to test whether automated judgments are sensitive to the evaluation setup and whether observed patterns generalize across configurations and judge models. Beyond aggregate agreement, we examine metric-level correlations, evaluator consistency, and systematic human-LLM disagreement to identify which conver- sational attributes can be judged reliably by automation and which remain sensitive to interpretation and context. Effective voice-agent evaluation is also shaped by pipeline-level factors such as speech generation, streaming, and error propagation across ASR, reasoning, and tool-calling stages, motivating our focus on comparing how human and LLM judges score the same interactions end to end. Our results show that LLM- based evaluation can serve as an effective component of large- scale voice-agent assessment, but that its reliability is metric- and configuration-dependent rather than uniform. This pro- vides an empirical framework for identifying which metrics suit automated evaluation and supports hybrid pipelines in which LLM judges handle scalable assessment while human evaluators remain engaged for metrics that demand contextual interpretation and higher-confidence judgment.

Agentic large language model (LLM) systems are reshaping scientific workflows in chemistry and drug discovery, but evaluating their open-ended, tool-augmented outputs remains a fundamental bottleneck. The LLM-as-a-Judge paradigm has emerged as a scalable alternative, but existing drug discovery benchmarks deploy LLM judges without validating their alignment with human experts. In this work, we present an LLM-as-a-Judge evaluation framework for ChatInvent, an agentic drug discovery assistant deployed at AstraZeneca, with five contributions. First, we define four output-quality evaluation dimensions---Completeness, Relevancy, Structural Clarity, and Scope Adherence---alongside deterministic Tool Call Correctness checks. Second, we validate the judge through a human alignment study with five expert annotators, comparing Gemini 3.1 Pro, Claude Opus 4.7, GPT-5, and Llama 3.1 70B as candidate judges. Third, we optimize the best-performing judge using few-shot demonstrations of human-annotated examples, improving alignment with the human majority vote from 0.80 to 0.86. Fourth, applying the optimized judge to 70 held-out questions, we surface concrete limitations and find no strong evidence that informal phrasing degrades output quality; it may, however, still be helpful to have the LLM rewrite the original question before querying the agent. Finally, we extend the framework to 38 adversarial questions that are ambiguous, invalid, out-of-scope or ethically sensitive, and show that the agent's refusal behavior is guided by the stated intent of a request. Our framework provides a reusable template for human-aligned evaluation of agentic systems in scientific domains.

LLM-as-a-Judge, which leverages a large language model to evaluate natural language generated by another AI application or model, has become a standard, scalable approach for accelerating and extending costly human evaluation. Yet most work treats a judge as a static artifact, evaluating it once at construction or against a fixed benchmark. We argue instead that an LLM judge operating in a deployed system is better understood as having a lifecycle. It must be built, trained, deployed, and continuously maintained as the surrounding data evolves, and each phase poses distinct technical and operational challenges. We present such a lifecycle for the LLM judges that evaluate recommendation explanations at Netflix. Everything we report comes out of a series of controlled online member-facing experiments, in which our pipeline generated and the judges assessed hundreds of thousands of distinct show-level explanations per week across a changing catalog. Our framework has four phases. (I) Birth defines the evaluation criteria and builds curated benchmark datasets with human labels and rationales. (II) Training refines the judges' rubrics via Reasoning-Aligned Rubric Tuning (RART), which uses a meta-judge over reasoning output as the learning signal. (III) Deployment puts one judge in two online roles, quality gating and reflective generation. (IV) Monitoring runs a continuous Human-in-the-Loop (HITL) alignment process that detects drift and triggers re-tuning behind a human review gate. We report results from a five-week online A/B test over tens of millions of members on the Netflix mobile app, in which judge-aligned explanations shifted member viewing toward novel content (previously unwatched) and increased successful browse-to-play sessions relative to a no-explanation control, with no quality-related escalations.

LLM judges have become central infrastructure for model evaluations, online grading, and reward modeling. Judges are typically validated by accuracy on golden data, but accuracy says little about whether they are stable under re-prompting, challenge, or sustained pushback. We introduce the \emph{Wiggle Framework}, a unified stress test for epistemic stability in LLM judges. The framework decomposes judge robustness along three dimensions: Mechanical Consistency (stability under re-prompting and reframing), Single-turn Conviction (stability under a single challenge), and Multi-turn Persistence (stability under sustained or adaptive pressure). We use the framework to study 9 frontier models across 14 judging tasks spanning safety, toxicity, AI writing detection, and political-response evaluation. Every model exhibits substantial wiggle as a judge --- flipping verdicts 25--71% of the time under static pushback, and 62--91% with an adversarial LLM persuader. Critically, we find that pressure that succeeds in changing a judge's verdict is almost always net-corrupting with respect to ground truth. Beyond the framework itself, we identify baseline jury majority strength as the most effective single-shot signal for anticipating which items wiggle. Taken together, this is the first apples-to-apples cross-dataset comparison of mechanical, conformity, and persuadability tests in a judging context.

Improving reasoning LLMs requires the ability to judge the quality of long reasoning traces for effective reasoning data curation, strong training signals during reinforcement learning, and an in-depth understanding of reasoning behaviors during model performance evaluation. Additionally, surfacing reasoning mistakes that the model makes would enable improving the model's performance at runtime through providing feedback. Due to the difficulty of this complex task on long reasoning traces, single-model judges (even frontier models) do not do well at identifying reasoning defects. Additionally, leveraging frontier models during online training of reasoning LLMs is generally prohibited due to guardrails in terms of use. In this work, we introduce Reasoning Jury, a system that replaces the single judge with a jury of LLMs and a moderated consensus mechanism, to improve the fidelity of judgments for identifying reasoning defects. In reasoning jury, defects of a reasoning trace and their severity are surfaced through a deliberation where a moderator conducts a discussion amongst the jury where the jurors critique each other's judgments and get to modify their initial votes. The moderator derives a consensus through deliberation amongst jurors or consolidation of judgements. We show that Reasoning Jury with a jury of open-weight models (e.g., gpt-oss-120b) is able to significantly outperform frontier models (opus-4.6, sonnet-4.6, and gemini-3.1-pro) at correctly identifying reasoning defects. Besides accuracy performance improvements, the aggregated cost of the jury (initial verdicts, deliberations, consolidation, etc.) is a fraction (8 to 15%) of the cost of running frontier models in LLM-as-a-judge setup. We also show how these judgements can be leveraged to understand failure modes of reasoning LLMs on benchmarks, which allows much deeper understanding of a model's performance.

Rubric-based evaluators commonly treat rubrics as prompt context or flat criteria: they specify what to judge but leave criterion composition implicit, even when natural-language rules state it. We introduce Graph-Structured Rubrics (GSR), which compiles a rubric into a response-independent typed evaluation graph before observing responses. Criterion nodes elicit judgments; transformation, reduction, and gating operators compose them through named ports; and a task-specific output mapping, termed Readout, converts the unique sink into a score or preference. Compilation rejects malformed or type-incompatible graphs. Pointwise evaluation judges rubric dimensions separately before graph aggregation; pairwise evaluation reuses the graph with one judgment for each candidate under every criterion. Under GPT-OSS-120B, GSR improves exact score agreement by 0.62--6.75 percentage points over Prometheus-style scoring on four pointwise datasets and achieves the numerically highest end-to-end pairwise accuracy on two preference benchmarks under native tie and abstention policies.

Generating actionable financial advice from business records demands that models integrate numerical reasoning, domain knowledge, and sound judgment, while avoiding recommendations that could harm the business. Direct supervision is difficult: historical decisions are not necessarily optimal, and high-quality free-form labels are expensive to obtain. We formulate financial advice generation as a reinforcement learning problem and fine-tune an open-weight language model using Group Relative Policy Optimization (GRPO). Our reward is an LLM-as-a-judge rubric that scores each recommendation across multiple binary dimensions of advice quality, augmented with a safety gate for harm prevention. Since LLM-based evaluation alone cannot confirm whether improvements reflect genuine business value rather than adaptation to the judge, we complement it with a judge-independent audit based on a standard doubly-robust Conditional Average Treatment Effect (CATE) estimator. Under this observational off-policy audit, our trained LLM achieves approximately twice the estimated gross-profit lift of the strongest evaluated commercial baseline ($0.0228$ vs.\ $0.0104$), together with the lowest downside rate and the least negative tail risk of any policy evaluated. Notably, the two evaluations do not rank the baselines identically: the untrained base model places last on the judge rubric but second on the causal audit, indicating that the audit captures a signal the judge does not. Our results demonstrate that GRPO with a finance-grounded reward signal can produce substantially more useful business recommendations than commercial LLMs, and that a judge-independent causal audit is a valuable complement to, rather than a confirmation of, LLM-as-a-judge assessment in financial NLP.

Mobile agent benchmarks increasingly rely on LLM-based judges to evaluate task completion, yet the reliability of these judges on mobile agent trajectories remains largely unexamined. We introduce MobileJudgeBench, a benchmark for systematically evaluating LLM-as-judge methods on mobile agent trajectories. Our benchmark comprises 931 human-annotated trajectories spanning 6 mobile agent benchmarks, 4 agent models, and 68 apps. We evaluate 6 judge methods (five adapted from SPA-Bench, A3 with two modes, AndroidArena, and AgentRewardBench, plus a simple baseline we design) across multiple LLM backends. Our experiments reveal three key findings. First, a simple baseline judge with sampled screenshots is competitive with, and often exceeds, purpose-built methods, indicating that more elaborate judge pipelines do not consistently improve judge quality; among competitive methods, the LLM backbone is the primary driver. Second, benchmark quality metrics reliably predict real-world judge utility: they correlate with both agent ranking fidelity for evaluation and downstream performance when judges serve as reward signals for on-policy reinforcement learning. Third, failure analysis across two LLM backends uncovers qualitatively opposite failure profiles, one conservative and the other permissive, linked to the backbone's precision-recall characteristics.

Analytic dashboards combine coordinated views and interactions for data exploration and decision-making. Recent models can generate them from data and natural-language goals, but evaluating their usefulness remains difficult. Dashboard generation is open-ended, and neither static appearance nor successful execution alone captures analytical support and interaction quality. We introduce DashArena, to our knowledge the first benchmark for open-ended, task-grounded generation of interactive analytic dashboards. Its key innovation is to require each system to generate both a dashboard and a replayable interaction trajectory. A browser executor replays the trajectory and turns the system's intended analytical workflow into reproducible visual and execution evidence. A VLM judge compares candidates using this evidence, and Bradley--Terry aggregation produces the leaderboard. We further distill the judge into the open-weight DashJudge-8B. Human evaluations show that DashJudge-8B effectively reproduces human judgments and ablations show that interaction evidence improves judge agreement. Experiments with frontier models reveal persistent rendering, analytical, and interaction failures. Together, these results show that realistic dashboard generation remains challenging and that interaction-aware evaluation captures failures missed by static or execution-only checks.

Large language models (LLMs) are increasingly used as relevance assessors in information retrieval (IR) evaluation, raising questions about how assessor framing affects judgment reliability and downstream system comparison. We study persona conditioning as a diagnostic mechanism for exposing LLM assessor sensitivity. Using task-oriented personas drawn from two complementary sources (PersonaHub and NVIDIA Nemotron-Personas-USA), we instantiate five assessor roles emphasizing intent interpretation, domain expertise, contrastive judgment, evidence verification, and global search-quality assessment, compared with a standard UMBRELA baseline. Across six LLM backbones on TREC DL20 and RAG24, our analyses reveal structured rather than uniform assessor sensitivity. Judgments usually remain close to the baseline while shifting assessment strictness, evidential threshold, or interpretation emphasis rather than producing widespread relevance reversals. At the system level, high-capacity models preserve system-ranking agreement, while smaller models amplify persona-induced instability. Local rank-displacement analysis shows sensitivity concentrates on particular retrieval systems and system types, especially neural ranking/reranking systems on DL20 and RAG-oriented pipelines on RAG24. Persona source matters less than assessor role and model capacity. These findings position persona-conditioned judging as a controlled sensitivity probe for stress-testing LLM-based IR evaluation pipelines and identifying systems whose evaluation outcomes are sensitive to assessor framing.

AI video generation has advanced rapidly and entered widespread commercial use. As a result, quality differences among videos produced by state-of-the-art AI video generation models~(AIVGMs) have become increasingly difficult to discern using conventional evaluation criteria, such as visual fidelity and semantic instruction following. Meanwhile, human evaluation now requires more expertise and sustained attention, substantially increasing annotation costs. This calls for automated evaluation that can reliably distinguish fine-grained differences among advanced AIVGMs with minimal human intervention. To address this challenge, we present RAVEN-Eval, a rubric-guided automated evaluation framework for AIVGMs, built primarily on the LMM-as-a-judge paradigm. Through an automatic task curation and quality-filtering pipeline, RAVEN-Eval curates 150 text-to-video~(T2V) tasks and 100 image-to-video~(I2V) tasks, and systematically collects more than 4,500 AIGVs. At its core, RAVEN-Eval adopts rubric-guided automated LMM preference judgement, in which LMM judges conduct pairwise comparisons according to task-specific rubrics. It further introduces an anchor-based model insertion approach to reduce the evaluation cost of incorporating new models. Finally, we evaluate 20 high-performance AIVGMs, as well as the judging capabilities of 13 LMM judges, and establish the RAVEN-Eval Leaderboards. Overall, RAVEN-Eval paves a scalable path for automatic and trustworthy evaluation of rapidly evolving AIVGMs.

We extend the DS@GT ARC working-note submission to the Touché 2025 Retrieval-Augmented Debate task. The task has two subtasks: generating the next utterance in a simulated debate, and evaluating debate responses according to the Gricean maxims of Quantity, Quality, Relation, and Manner. The DS@GT ARC submission consisted of six leading LLMs from three providers through a retrieval-augmented prompting pipeline. We summarize the results from the working paper and explore whether multi-LLM evaluator agreement is a reliable proxy for official evaluation performance. The analysis shows that frontier LLM systems are strong response generators, and as evaluators they agree strongly within model families. However this consensus does not reliably track the official evaluation target, with the largest gap on the Quality maxim. The accompanying source code for this paper is located at https://github.com/dsgt-arc/touche-2025-rad and https://github.com/dsgt-arc/touche-2025-rad-analysis.

Language-model judges and reward models enable scalable supervision, but finite optimization can exploit evaluator errors rather than improve response quality. We characterize this failure through the covariance geometry of evaluator ensembles. For calibrated judges, the ensemble mean retains common-mode error along the all-ones direction, whereas cross-judge disagreement captures only orthogonal error. Consequently, disagreement can be high despite robust aggregation, or low while shared response-dependent errors persist. We prove that common-mode error is not identifiable from internal judge scores alone. Under a joint sub-Gaussian model, we bound best-of-K selection overstatement and target-quality regret, extending the guarantees to predictably adaptive search under conditional calibration. The resulting search terms scale as the square root of log K and are asymptotically tight for Gaussian projected errors. We further show that noisy quality proxies introduce artificial rank-one covariance without changing disagreement, and propose a bounded two-anchor Bernstein certificate for finite-search error and regret. Fixed-seed Gaussian stress tests over 120 (J, rho, K) configurations and real-model audits validate the theory while revealing the limits of disagreement-based diagnostics under increasing search pressure.

Multimodal LLM judge panels can cross-reference peers, but a quoted peer judgment may itself be untrusted. We expose source-blind anchoring as a text-level attack surface in vision-language model (VLM) panels. Quoting independent visual judgments creates large anchoring gaps (19--26 percentage points) under both self and peer framing. A matched-content, label-only control changes the broken rate by only $-0.17$pp (95% CI $[-0.68,0.35]$), showing that the self/peer label itself does not explain the effect. Under our tested construction, deliberately generated, concise wrong quotes overturn originally-correct verdicts 1.5--2.7$\times$ more often than naturally occurring wrong peer statements, with bootstrap 95% CIs excluding parity across two datasets and seven VLM judges. Because the two statement populations differ in selection and form, this ratio measures differential damage under the tested attack rather than a provenance-only causal effect. We then introduce panel-consensus verification, which cross-checks a quote against independently collected blind votes. It blocks 84.9% of fabricated attacks, cuts their net harm by 97.5%, and preserves the positive but statistically inconclusive point estimate for genuine peer information under leave-one-out re-verification. These results identify a low-cost attack surface and a concrete defense for safer multimodal collaborative evaluation.

An LLM judge deployed inside a reasoning pipeline does not merely measure quality, it decides which answer ships. We show that the cost of that decision depends less on judge accuracy than on the decision rule the judge is embedded in. On frozen candidate pools from four GRPO policies, an unconstrained scalar DeepSeek-R1-7B judge buys almost nothing over answer-level majority vote (+1.0 pp on 500 GSM8K questions, +0.34 EM on 300 HotpotQA questions), and on a frozen-rule 30-question confirmation split it is 10 points worse than majority, a judge that destroys accuracy while scoring candidates confidently. We then subordinate the same judge to Evidence-Locked Derive-Gate-Repair (EL-DGR), a task-adaptive non-compensatory rule under which a judge preference may override evidence-supported consensus only with an extractive evidence certificate, and a repair only when neither alternative is certified and the repair is. With no change to the judge, the candidates, or the budget, EL-DGR reaches 58.2% on GSM8K (vs. 56.8% judge, 55.8% majority, 55.4% first candidate) and 17.33 EM / 25.46 F1 on HotpotQA (vs. 15.67/23.49, 15.33/23.19, 15.33/22.97), improving on first-candidate GRPO by +2.8 pp (exact McNemar p=0.0026) and +2.00 EM (p=0.070, borderline). A decision audit shows why: EL-DGR overturns consensus on only 8 of 30 pilot questions and never converts a correct consensus into an incorrect answer. We also report what did not work: the same seven-channel decomposition used as a step-level gated training reward is null, and corrected channel-drop ablations show no channel is individually necessary (p=1.0 throughout). The practitioner-facing finding is negative about judges and positive about admissibility, bound the judge's blast radius rather than trying to make it accurate.

LLM benchmarks can build an organization's reputation and attract customers, but only when results are transparent and verifiable. Unverified claims that DeepSeek R1 outperformed OpenAI's o1 contributed to market panic on January 27, 2025, when Nvidia lost USD589 billion in market value. Yet vendor benchmarks often depend on an honor system. Academic reassessments and independent leaderboards have found undisclosed changes to proprietary models, contaminated training data, and selective reporting. LLM-as-a-judge methods scale evaluation by reducing human review. Studies, however, suggest that judges may show identity-aware bias, scoring an answer according to its source model rather than its quality. This bias has not been fully measured or corrected across politically sensitive, reasoning-intensive, and preference-based tasks. We examine this problem using seven verifier models: GPT-OSS 120B, Llama 3.3 70B, GLM 5.1, Qwen3 32B, DeepSeek V4 Pro, Mistral Large3, and Sarvam M. They score anonymous and identity-disclosed responses from three primary models on 58 factual, reasoning, political, and preference-based questions. Identity disclosure slightly raises scores for factual questions, moderately affects stress-reasoning tasks, and causes large changes for geopolitically sensitive topics. Notable results include GLM5.1 (+7.00 points, p = 0.0249) and Llama 3.3 70B (+1.56 points, p = 0.00). We also introduce a blockchain-based commit-reveal protocol using Autonomous Economic Agents on an Ethereum-compatible ledger. In Phase 1, each judge records a one-way hash of its score and a secret salt before candidate identities are revealed. In Phase 2, the identity and raw score are disclosed and verified on-chain. This creates a tamper-evident audit trail that separates blind evaluation from post-hoc claims and reduces the verification burden on independent researchers and leaderboard operators.

The rapid advancement of large language models has transformed survey writing from a months-long manual effort into an automated process. As generation scales, reliable evaluation becomes the bottleneck, and LLMs are increasingly used as survey evaluators. However, existing approaches largely rely on off-the-shelf LLM-as-a-judge methods without systematic alignment to human reviewers, and there remains a lack of systematic frameworks for quantifying alignment with human reviewers. To address this gap, we propose SurveyReview, a reviewer-aligned, multi-dimensional benchmark and dataset for survey evaluation. We collect and annotate 675 survey papers with 1,630 review reports. We structure authentic peer-review reports by converting free-form comments into four-dimensional scores (Readability, Criticalness, Comprehensiveness, Structure) paired with supporting rationales. We further release standardized train/test splits and an evaluation protocol to measure alignment between automatic evaluators and human reviewers. To validate the benchmark, we develop SurveyAlign, a strong baseline evaluator by fine-tuning Qwen3-32B with LoRA on our annotated data, augmented with external knowledge for knowledge-intensive dimensions. On the test set, SurveyAlign substantially improves reviewer alignment over prompt-based judging with GPT-5.2, reducing average MSE from 2.28 to 1.38 and MAE from 1.15 to 0.69 across all four dimensions. Our contributions are twofold: (1) we establish the first multi-dimensional, reviewer-aligned dataset with a reproducible evaluation framework for survey reviewing; (2) we develop a strong baseline evaluator that substantially improves alignment with human reviewers, providing a competitive reference for future research. Our code and data are available at https://surveyreview.github.io

Human-Oriented Binary Reverse Engineering (HOBRE) aims to transform decompiled pseudocode into a more human-friendly representation, thereby reducing the cognitive burden of reverse analysis and improving efficiency. However, reliably evaluating HOBRE outputs remains a fundamental challenge: human evaluation is costly, time-consuming, and difficult to scale, while existing automated metrics either require executable test cases and runtime environments that are often unavailable for real-world binaries, or rely on high-quality source code references that are typically inaccessible and fail to capture semantically equivalent but lexically diverse outputs. Although LLM-as-a-Judge paradigm is naturally well-suited to HOBRE evaluation, its effectiveness remains underexplored. This paper presents the first systematic investigation of the LLM-as-a-Judge paradigm for HOBRE across three representative tasks: function name recovery, binary code summarization, and decompilation optimization. We introduce BinJudgeBench, the first expert-annotated, reference-free evaluation benchmark based on multi-dimensional human judgment, where LLM-as-a-Judge achieves an average correlation of 63.20% with human judgment, outperforming traditional automated metrics at 35.04%. By analyzing judge configurations across backbone LLMs, prompting strategies, and decoding temperatures, we find that no ``one-size-fits-all'' configuration exists, as the optimal setup varies across tasks and individual samples. To address this, we propose BinJudge, which employs a lightweight routing mechanism to adaptively select the optimal judge configuration for each task and sample. BinJudge improves correlation with human experts by 4.5%-24.7% and reduces API cost to 0.06$\times$-0.84$\times$ of that of static best configurations, providing a scalable, cost-effective, and high-fidelity automated evaluation scheme for HOBRE.

LLM judge panels are a standard evaluation tool, but prior work reports highly correlated panel errors: nine judges provide roughly the effective information of two independent ones, and aggregation closes only a small fraction of the gap. A natural remedy--a signal from a different evidence source, e.g., executing a test suite--produced no distinguishable change in the panel's effective-vote count at scale (-0.04, 95% CI [-0.10, +0.02]). Aggregate dependence and conditional decision utility are different questions. Elementary majority arithmetic fixes the affected set for single-ballot substitution: only decisions with a one-vote margin can change. The empirical question is whether panel error rates rise and useful substitutions concentrate there. They do: the entire accuracy gain concentrates on these pivotal queries, where it is large (+10.4 to +23.3 percentage points across three headline configurations), and is exactly zero elsewhere. We confirm the pattern across three code benchmarks and four panel sizes (a 9-judge extension and 56 dependent subsampling checks, gain +6.5 to +16.1 percentage points). On HumanEval+/MBPP+, a majority-side replacement rule raises overall accuracy from 82.44% to 85.62% while invoking the signal on 16.2% of queries; signal-only remains stronger at 87.60%. Thus population-level dependence diagnostics and margin-stratified utility are complementary, and the affected-set characterization yields a call-reduction rule for any specified single-ballot substitution policy.

Large Language Models (LLMs) are often used as evaluators of text quality, known as LLM-as-a-Judge, which can outperform conventional automatic evaluation metrics that rely on reference texts. However, LLM evaluators tend to generate particular scores regardless of the context of the evaluated text, which is known as scoring bias. This study proposes a novel method to mitigate this scoring bias. An LLM is instructed to randomly generate number tokens, and the latent numerical bias of the LLM is identified by measuring the deviation of the observed distribution of numbers from the uniform distribution. A definition of a downstream task, for which an LLM evaluator is used, is added to the prompts for random number generation to measure task-specific latent number bias. In the evaluation by an LLM, the token generation probabilities for a given input are rectified considering the LLM's latent number bias. Results of the experiment on four different tasks, evaluation of LLM alignment, evaluation of summarization, Semantic Textual Similarity, and Semantic Textual Relatedness, demonstrate that our proposed method outperforms the baselines, including an LLM without debiasing and previous calibration methods. In addition, it is confirmed that scoring bias varies across LLMs, tasks, and score ranges, indicating the importance of measuring latent number bias as the case may be.

LLM agents increasingly execute long-horizon tasks through tool use and environment interaction, shifting evaluation from final-response scoring to verification of complete executions. For skill-augmented agents, verification additionally requires the procedural knowledge encoded in task-time skills, because this knowledge indicates what evidence to inspect and which failures are task-critical. However, existing judge benchmarks often expose final responses or static trajectories, and rarely combine task-time skills with directly inspectable artifacts and environments. We therefore introduce SkillTV-Bench, a 681-case benchmark of real agent trajectories from 50 tasks across eleven domains, designed to evaluate skill-aware trajectory verification for both LLM-as-a-Judge and Agent-as-a-Judge methods. Additionally, we propose SkillTV-Evolve, which externalizes verification knowledge as a reusable JudgeSkill that guides an agent judge to plan targeted inspections and issue evidence-grounded verdicts. On a disjoint development pool, an automated evolution loop further refines the JudgeSkill using misjudged cases. On SkillTV-Bench, the refined skill increases the same agent judge's accuracy by 14.8 percentage points. In offline rollout-pool selection, it increases selected-trajectory success from 22.9% with one rollout to 45.5% with ten rollouts. The code and data are available at https://github.com/HanZhi306/SkillTV-Bench