Daily Papers

Recent vision-and-language navigation (VLN) policies act on their visual history, and an incorrect action can take them off the instructed route. Paired interventions on four VLN policies show that clearing the history doubles pooled success at the onset of a deviation but lowers it on route, while later in a deviation, removing only the off-route frames helps much less. After a deviation, the retained history keeps the policy from turning where a turn is needed, which we call behavioral inertia. We propose MOON, a framework that equips a frozen VLN policy with two lightweight modules, adding less than 1% of its parameters. A deviation judge decides when to forget, raising an alarm that clears the history when it predicts that the agent is off route. A recovery policy, trained with GRPO from alarm states, learns how to move on, turning more of the successes found only by sampling into a single greedy rollout. Experiments on R2R-CE and RxR-CE show that MOON reaches the highest R2R-CE success rate among the compared methods and raises the success rate of its frozen base by 5.2 and 6.0 points, and an ablation with a random trigger indicates that this gain depends on when the history is cleared. Without retraining, MOON also improves zero-shot object-goal navigation on HM3D-OVON. Together, these results suggest that deviation recovery depends on knowing when to forget, not only how to act. Code is available at https://github.com/AutoLab-SAI-SJTU/MOON.

Multimodal Large Language Models (MLLMs) have been increasingly used as automatic evaluators-a paradigm known as MLLM-as-a-Judge. However, their reliability and vulnerabilities to biases remain underexplored. We find that many MLLM judges fail to reliably integrate key visual or textual cues, yielding unreliable evaluations when evidence is missing or mismatched, and exhibiting instability under semantically irrelevant perturbations. To address this, we systematically define Compositional Bias in MLLM-as-a-Judge systems and introduce MM-JudgeBias, a benchmark for evaluating it. MM-JudgeBias introduces controlled perturbations across Query, Image, and Response, and evaluates model behavior via two complementary metrics: Bias-Deviation (BD) for sensitivity and Bias-Conformity (BC) for stability. Our dataset of over 1,800 curated and refined multimodal samples, drawn from 29 source benchmarks, enables a fine-grained diagnosis of nine bias types across diverse tasks and domains. Experiments on 26 state-of-the-art MLLMs reveal systematic modality neglect and asymmetric evaluation tendencies, underscoring the need for more reliable judges.

Modern LLM agents combine long-term memory for personalization with tool-calling interfaces for taking actions in the world -- a combination underpinning contemporary production systems. We study a previously unexamined failure of this combination: when personality-driven biases stored in memory (cost-consciousness, impatience, risk tolerance, etc.) silently affect tool calls in contexts where they are not applicable. We call this memory-induced tool-drift and operationalize it through MEMDRIFT, a benchmark of 105 scenarios spanning five bias dimensions and seven professional domains, generated through an automated adversarial pipeline. Across seven frontier models -- including those with extended reasoning -- biased memories raise deflection scores (a judge-scored measure of parameter deviation from unbiased baselines) by up to +3.6 points on a 1--5 scale. Tool-drift persists when memory management is handled by three production memory architectures. The phenomenon affects real-world tools: scanning 6{,}062 tools across 288 verified MCP servers, we flag 608 with susceptible parameters and confirm tool-drift on a validated subset. Mechanistically, biased memories act as implicit steering vectors, pushing activations along the same latent directions as explicit behavioral instructions. They also redistribute attention from task-relevant context toward memory entries with surface-level keyword overlap to the target parameter. Standard defenses -- prompt-based relevance instructions and memory filters -- reduce drift but do not eliminate it. As agents take increasingly consequential actions on a user's behalf, memory-induced tool-drift represents a systematic vulnerability that current safeguards do not address, motivating dedicated defenses at the intersection of memory management and tool-call generation.

Continuous evaluation of LLM products relies on a strong LLM judge treated as ground truth: a cheap monitor scores every interaction and a team is paged when the score drifts down. But the judge is itself a model behind an API, and a silent version bump or scoring-prompt update changes how it scores -- so every drift alarm is ambiguous between a worse product and a changed judge. We resolve the ambiguity with a fixed, human-labeled anchor set that the current judge re-scores at a steady interleave, a second betting e-process on the judge-versus-human gap, and a guard-window rule returning a verdict in {none, system, judge}. We prove anytime-validity, one-way identification (only the judge can move the anchors), an attribution race whose design law is that the anchors must out-run the main process they guard, and process orthogonality. On two real judge changes, a silent version bump is detected as judge drift in 60/60 runs with zero judge-to-system misattribution, and a contaminating strict-prompt change is correctly attributed on 110 of 120 runs at guard width 300 -- while the industry-default rolling z-test false-alarms on 75% of drift-free streams. Every experiment replicates on a second domain (TL;DR summarization) with nothing re-tuned, and where the domains differ the differences are the ones the race predicts: the strict-prompt change shifts scores harder there, so the anchors fire faster and attribution becomes perfect (240/240). The monitor runs at approximately 0.64 of the cost of strong-judging every item, or 0.21 in a cheaper-but-deafer regime.

Ambient AI scribes draft clinical notes, and published audits find their dominant error is omission: information the encounter established that the note fails to record. The standard check is an LLM judge: a second model reads the note against the transcript and flags problems. We ask whether judges detect omissions. Public corpora cannot supply the answer key: their clinician reference notes and transcripts are materially discrepant. Our benchmark has 500 single-error note pairs from audited fact sheets, 298 with a named fact certainly absent and 202 added-or-altered controls. Across eight judge designs, paired discrimination (the flawed note below its clean twin, 0.5 a coin flip) reads 0.79-0.94 on added or altered content and 0.50-0.63 on omissions. On single notes, no design flags omissions reliably more often than perfect notes. Wording changes, voting and GEPA prompt optimisation move the operating point without creating usable detection. Restructuring the task recovers it: list the facts the transcript establishes, then check the note for each. Two methods reach it independently and trade off: a per-fact pipeline, and a GEPA-evolved prompt doing the same in one call. The pipeline's flags name the missing fact and its severity at 2.7% false alarms. The single call detects more (36.9% against 24.6%, p=0.002) at 6.2% false alarms and a tenth of the cost per note. A physician author validated 70 items and, where the two routes disagree, sided with the pipeline on 10 of 10 (p=0.002). A second clinician, not an author, graded the severity rubric blind and agrees to within a grade. On real vendor notes from a companion census no benchmark threshold transfers, but the re-calibrated single call detects more than the best of the eight at half its false-alarm rate. Omissions whose fact is restated elsewhere defeat both routes. We release the benchmark, prompts and judgements.

Safety benchmarks such as HarmBench rely on LLM judges to classify model responses as harmful or safe, yet the judge configuration, namely the combination of judge model and judge prompt, is typically treated as a fixed implementation detail. We show this assumption is problematic. Using a 2 x 2 x 3 factorial design, we construct 12 judge prompt variants along two axes, evaluation structure and instruction framing, and apply them using a single judge model, Claude Sonnet 4-6, producing 28,812 judgments over six target models and 400 HarmBench behaviors. We find that prompt wording alone, holding the judge model fixed, shifts measured harmful-response rates by up to 24.2 percentage points, with even within-condition surface rewording causing swings of up to 20.1 percentage points. Model safety rankings are moderately unstable, with mean Kendall tau = 0.89, and category-level sensitivity ranges from 39.6 percentage points for copyright to 0 percentage points for harassment. A supplementary multi-judge experiment using three judge models shows that judge-model choice adds further variance. Our results demonstrate that judge prompt wording is a substantial, previously under-examined source of measurement variance in safety benchmarking.

Automated ideation systems are often evaluated on the novelty of the ideas they produce, and that judgment is increasingly delegated to large language models. Such judges are typically built ad hoc and validated, if at all, on human-authored papers rather than on the generated ideas they are meant to score. So, how do novelty judges perform? Not well. We present a systematic controlled study of novelty evaluation design choices. We first build an evaluation set automatically, mining OpenReview for passages where reviewers explicitly affirm or dispute a paper's originality and keeping only submissions with unanimous agreement at the extremes of their research area; we pair these with ideas from a vanilla LLM generator. Across six judges, we find that small prompt design choices have large consequences; e.g., simply telling the judge that reviewers found one idea novel and the other not can change its verdict on more than half of the identical idea pairs it is shown, shifting pairwise accuracy by over 50 points and occasionally pushing it below chance. The same change helps one judge and hurts another. Retrieval and larger reasoning budgets help little, and two purpose-built novelty evaluators are outperformed by our cheapest prompted baseline. These results raise questions about reported novelty gains of automated ideation systems, and call for robust novelty evaluation methods.

Computer-use agents are capable of completing complex tasks, increasing the use of automatic judges to determine success, either for training or for evaluation without human involvement. Despite their flexibility, their reliability on long tasks spanning multiple applications remains unclear. A trajectory, composed of long sequences of screenshots and actions, may appear complete, but in reality violates constraints from the instruction or introduces an unwanted side effect. To identify these errors, a judge needs to examine the trajectory with respect to the user's instruction. To this end, we introduce AgentHorizon, a benchmark of 1,373 computer-use tasks (instruction-trajectory pairs) drawn from 166 hours of human-recorded trajectories spanning three operating systems. By recording trajectories for closely related instructions, we can construct negative tasks by swapping the instructions. This paired design evaluates judges on their ability to distinguish a successful trajectory from one that completed a similar (but incompatible) request. We release the benchmark under three splits: a frontier split, AgentHorizon (AH), a simplified split, AgentHorizon-Simple (AH-S), and a development split, AgentHorizon-Development (AH-D). We further evaluate eleven judges by (1) directly passing the full trajectory (with up to 300 screenshots and actions), and (2) using them as coding agents across five agent harnesses. We find that our best agentic judge, GPT-5.5, achieves 80.9% balanced accuracy on the AH subset. We find that tool-use improves certain models but results in worse performance for open-weight models, and that judges differ drastically in their ability to accept a valid trajectory and reject failed ones. Our findings highlight the need for judges that are capable of locating and verifying often hidden evidence that a task was properly completed inside long interaction histories.

In a legal system, judgment consistency is regarded as one of the most important manifestations of fairness. However, due to the complexity of factual elements that impact sentencing in real-world scenarios, few works have been done on quantitatively measuring judgment consistency towards real-world data. In this paper, we propose an evaluation metric for judgment inconsistency, Legal Inconsistency Coefficient (LInCo), which aims to evaluate inconsistency between data groups divided by specific features (e.g., gender, region, race). We propose to simulate judges from different groups with legal judgment prediction (LJP) models and measure the judicial inconsistency with the disagreement of the judgment results given by LJP models trained on different groups. Experimental results on the synthetic data verify the effectiveness of LInCo. We further employ LInCo to explore the inconsistency in real cases and come to the following observations: (1) Both regional and gender inconsistency exist in the legal system, but gender inconsistency is much less than regional inconsistency; (2) The level of regional inconsistency varies little across different time periods; (3) In general, judicial inconsistency is negatively correlated with the severity of the criminal charges. Besides, we use LInCo to evaluate the performance of several de-bias methods, such as adversarial learning, and find that these mechanisms can effectively help LJP models to avoid suffering from data bias.

LLM-as-a-judge enables evaluation across diverse tasks, but inference cost and confidence reliability become critical at scale. We study whether a decision-only judge can provide an economical first pass and identify when stronger evaluation is needed. Comparing jev-as-a-judge with sixteen generative and reward-model judges, with blinded human adjudication, we find it within three percentage points of a state-of-the-art LLM judge, our strongest comparator, on ordinary preference and evidence-grounded factuality at 0.36% of the comparator's fee. Larger gaps arise when judgments require checking a derivation or resisting an elaborately written wrong answer. On several benchmarks, JEV's gap to this comparator is concentrated in low-confidence decisions. A frozen cascade that accepts confident verdicts and escalates uncertain ones retains 99% of the comparator's accuracy at lower cost.

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

The use of LLMs as automated judges ("LLM-as-a-judge") is now widespread, yet standard judges suffer from a multitude of reliability issues. To address these challenges, we introduce Verdict, an open-source library for scaling judge-time compute to enhance the accuracy, reliability, and interpretability of automated evaluators. Verdict leverages the composition of modular reasoning units (such as verification, debate, and aggregation) and increased inference-time compute to improve LLM judge quality. Across a variety of challenging tasks such as content moderation, fact-checking, and hallucination detection, Verdict judges achieves performance competitive with orders-of-magnitude larger fine-tuned judges, prompted judges, and reasoning models. Our framework establishes a foundation for scalable, interpretable, and reliable LLM-based evaluation systems for both researchers and practitioners.

Multi-constraint instruction following requires verifying whether a response satisfies multiple individual requirements, yet LLM judges are often assessed only through overall-response judgments. We introduce MCJudgeBench, a benchmark for constraint-level judge evaluation in multi-constraint instruction following. Each instance includes an instruction, a candidate response, an explicit constraint list, per-constraint gold labels in {yes, partial, no}, and controlled response-side perturbations. The evaluation protocol further includes evaluation prompt variants to test judge stability. We evaluate proprietary and open-source LLM judges using both correctness and inconsistency metrics, distinguishing intrinsic inconsistency under stochastic decoding from procedural inconsistency under prompt and response perturbations. Our results show that judge reliability has multiple dimensions: strong overall performance does not guarantee equally reliable detection across label categories, especially for rarer partial and no cases. Judges with higher correctness do not always have lower inconsistency. Evaluation with reasoning improves correctness but does not uniformly improve stability. These findings motivate evaluating LLM judges at the constraint level to study these failure modes.

LLM judges are widely used to evaluate agentic tool-calling systems, yet their reliability on structured, dependency-driven workflows remains largely unexamined. We present AgentJudgeBench, the first benchmark to systematically study LLM-as-a-judge reliability for agentic tool-calling over workflow DAGs, as distinct from the broader LLM-as-a-judge task of open-ended text or preference evaluation. The benchmark comprises 3,808 instances spanning six DAG topologies and three difficulty tiers, evaluated with five generators (3B-70B open-weight models and GPT-5.4) and six judges (20B to frontier scale) under paired with- and without-ground-truth conditions. Judge alignment degrades monotonically with task difficulty, 1.5x faster without ground truth, and on hard queries without ground truth all six judges converge to a narrow 77-82% band regardless of scale, revealing a structural ceiling driven primarily by task difficulty, though its height is partly prompt-dependent for weaker generators, that model capacity alone cannot overcome. Ground-truth exposure is not uniformly beneficial: it reduces alignment for GPT-5.4 (1.5 pp) and Gemini-2.5-Pro (3.9 pp), consistent with over-anchoring. Among mitigation strategies, chain-of-thought reasoning and judge temperature both have negligible effect, while structured evaluation rubrics improve alignment by up to 6.5 pp but do not generalize uniformly across judge-generator pairs. With ground truth, QwQ-32B best matches the programmatic reference, while a human validation study identifies GPT-OSS-120B as the most human-aligned judge; without it, frontier judges lead only marginally within the shared ceiling. These results expose fundamental limitations of current LLM judges and yield practical guidelines for reliable evaluation in agentic systems.

Almost every paper on LLM jailbreaks and prompt injection reports an attack-success rate (ASR), and that number is assigned not by people but by an automated judge: either a safety classifier trained for the task, or a general chat model prompted to grade. The judge is rarely checked. We check it. Using 596 human-labeled completions from the HarmBench classifier validation set, we compare the two judge families against human majority votes and then attack them. The two families fail in opposite ways. The dedicated classifier over-flags (precision 0.835, recall 0.974); three different LLM-as-judges keep high precision (0.81 to 0.94) but show erratic recall (0.06 to 0.65), so the same responses produce very different ASR depending on which judge scores them. The two families also differ sharply in robustness. Wrappers that leave the harmful text untouched and only add benign framing flip every LLM-judge between 57% and 100% of the time, and a single prepended refusal sentence accounts for much of this (39% to 88%). The dedicated classifier resists these surface attacks (at most 6.7%), but a white-box GCG attack on its open weights flips 70% of confident true positives (21 of 30; 95% CI 54 to 86%) even at a small optimization budget. A two-annotator audit confirms the attacks leave the harm intact: every one of 80 sampled flips still contained the harmful content. Because a large and growing share of reported ASR comes from LLM-judges, many such numbers are unreliable both on average and under deliberate pressure. We recommend that papers report judge precision and recall on a human-labeled slice, report ASR corrected for judge precision, and include an adversarial check of the judge. Our code is released.

Large language models are increasingly deployed as automated judges for evaluating other models, yet the stability of their verdicts under semantically equivalent prompt paraphrases remains unmeasured. We introduce JudgeSense, a framework and benchmark for quantifying this property via the Judge Sensitivity Score (JSS), defined as the fraction of paraphrase pairs on which a judge returns an identical decision. Evaluating nine judge models on 494 validated paraphrase pairs, we find that coherence is the only task where judges meaningfully differ, with JSS ranging from 0.389 to 0.992. On factuality, all judges cluster near JSS about 0.63, driven by a polarity-inverted prompt artifact; after correction, factuality JSS rises to about 0.9. Pairwise tasks (preference and relevance) exhibit degenerate always-A behavior in 8 of 9 judges, indicating strong position bias. Model scale does not predict consistency. We release code, decision logs, and a validated paraphrase dataset to support standardized JSS reporting.

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

Reinforcement learning against rubrics, lists of criteria graded by an LLM judge, has become a standard way to post-train language models on tasks with no deterministic answer. The rubric, however, is a fixed proxy for quality, never a complete description of it, and a policy trained against it long enough will learn to exploit the difference. We measure this directly. Training Qwen3-8B with Group Relative Policy Optimization (GRPO) on medical and science rubrics and grading out-of-distribution (OOD) benchmarks with both the training judge and a stronger gold judge, we find that the two scores diverge during training. The training judge's score keeps climbing while the gold judge's score peaks and then falls, by 3 points on HealthBench-Hard and by 22 points on ResearchQA. A judge with a fixed bias would shift the gold curve by a constant, not send it down while the training score rises, so the divergence is reward hacking, not judge noise. We propose Rubric Dropout, a one-line fix borrowed from neuron dropout. At every step, we randomly drop a subset of the rubric's criteria before computing the reward, so the policy never optimizes the same rubric twice. The dropped subset is shared across each rollout group, so GRPO's group-relative advantages stay comparable, and evaluation always uses the full rubric. Comparing no dropout against dropout at 30% and 50% on both benchmark pairs, dropout raises the OOD gold score at every matched checkpoint (+1 to +2 points on HealthBench-Hard, +6 to +7 points on ResearchQA), lowers the two hacking measures we track, and costs nothing in domain. Sweeping the dropout fraction shows a broad 30-50% sweet spot, while the natural alternative, reweighting criteria by how useful they are to training, performs worse than no intervention at all in our setting.

We assess indirect prompt injection in DeepSeek Harness (DSH), using AI-Infra-Guard (A.I.G) to construct tests, deliver controlled taint, execute DSH, collect traces, and judge outcomes. The study covers 14,560 controlled executions over 16 indirect-content channels, text and file carrier modes, 35 payload objectives, one unmodified baseline, and 12 attack methods. The experiment preserves DSH's agent loop, tool registry, model adapter, and session-event path; source tools and sensitive sinks are local fixtures, so attempted actions are recorded without external side effects. We evaluate each trace with a deterministic rule-based judge, (RuleJudge), and a semantic LLM-based judge, (LLMJudge). The strongest observed attack success rates are 17.0% under for fake-completion attack in text mode, 25.5% under for hidden Unicode in file mode, and 16.0% under for the skills channel in file mode. also assigns partial compliance more often than (7.3% versus 2.0%). We relate these results to DSH's treatment of tool results, additional contexts, and tool-call policy hooks, then identify controls that should sit between untrusted content and sensitive actions. Our code is available at https://github.com/Tencent/AI-Infra-Guard/tree/main/Research/deepseek-harness-security-assessment .

LLM-as-judge is now a standard tool for scalable evaluation, but judge performance is still often summarized by a single accuracy number. This aggregate view hides the deployment conditions under which a judge succeeds or fails. We introduce Conditional Accuracy Profiling (CAP), a post-hoc diagnostic framework that decomposes pairwise LLM-judge accuracy into eight conditions organized into content sensitivity, robustness, and rationale quality. CAP is benchmark-agnostic: it can be applied directly when a benchmark provides the required annotations, approximately through task-subset proxies, or through controlled augmentation when perturbation pairs can be generated. We instantiate CAP on seven LLM judges across six pairwise judging benchmarks, including judgerEva-Standard, a controlled testbed we created to support all eight conditions. CAP exposes profile differences hidden by aggregate accuracy: on judgerEva's judge-independent Hard-Constructed subset, the two judges most sensitive to omitted qualifications rank in the bottom three of seven by overall accuracy, so omission sensitivity is not predicted by aggregate accuracy. Across benchmarks, Position Robustness shows the strongest rank stability (mean Spearman barρ{=}0.87) but is itself fragile under JudgeBench-Pro adversarial stress, showing the largest mean accuracy drop among the shared conditions, though the dominant degradation channel varies by judge. Condition-level profiles provide a more actionable basis than aggregate accuracy for selecting LLM judges.

A sound judgment applies the law to established facts and weighs the circumstances in which they arose. However, existing methods swing between rigid statute matching and ungrounded discretion, benchmarks score a label or a rubric, and the experience that would supply the balance stays unverified. We formalize legal judgment as a reference-anchored task, whose object is a single decision that stays tied to the statute and to the circumstances at once. We introduce JusticeAxis, 256 real-world criminal cases from 18 countries with audio, image, and text evidence, and three lawyer-written judgments for every case: the recorded one and one for each failure. We further propose JusticeAgent, a harness whose element agents establish the facts and whose judge agent applies the law under skills carrying experience of the circumstances. Skills are distilled from execution trajectories and admitted only under Bayesian credible bounds. Experiments show that failure turns direction with scale: open-weight backbones drift to unsupported grounds, frontier models to the statutory default. We further verify that JusticeAgent, as a simple yet effective plugin, carries a frozen open-weight backbone to commercial level. Project resources are available at https://github.com/beita6969/JusticeAxis.

Deep research agents increasingly automate complex information-seeking tasks, producing evidence-grounded reports via multi-step reasoning, tool use, and synthesis. Their growing role demands scalable, reliable evaluation, positioning LLM-as-judge as a supervision paradigm for assessing factual accuracy, evidence use, and reasoning quality. Yet the reliability of these judges for deep research agents remains poorly understood, posing a critical meta-evaluation problem: before deploying LLM judges to supervise research agents, we must first evaluate the judges themselves. Existing meta-evaluations fall short in two ways: (1) reliance on coarse, subjective human-preference agreement; (2) focus on instruction-following or verifiable tasks, leaving open-ended agent executions unexplored. To address these gaps, we introduce REFLECT (REliable Fine-grained LLM judge Evaluation via Controlled inTervention), a meta-evaluation benchmark targeting fine-grained failure detection in agentic environments. REFLECT defines a detailed taxonomy of process- and outcome-level failure modes, instantiated by performing controlled and localized interventions on quality-screened agent execution traces. This yields verifiable, comprehensive, and fine-grained instances for validating the judge models. Our experiments show that current LLM judges remain unreliable: even the best-performing models achieve overall accuracies below 55% across reasoning, tool-use, and report-quality failures, with especially poor performance on evidence verification. Together, our taxonomy and findings expose systematic judge limitations, reveal tradeoffs in cost and reliability, and offer actionable guidance for building more reliable evaluation pipelines for deep research agents.

Self-improving policies continually expose new failure patterns, changing what their judges must be able to verify. However, current fixed judges constrain both optimization feedback and the discovery of useful training examples, limiting further self-improvement. This challenge is even more acute in embodied reasoning, where reliable evaluation must account for spatial grounding, causal reasoning, and safety-aware decision-making. We introduce VeriFine, an agent harness framework that scales verification through the co-evolution of the policy, training curriculum, and judge. The Policy Improvement Loop uses a rubric judge to diagnose recurring failures, construct an adaptive curriculum, and optimize the policy. When progress plateaus and verification becomes a bottleneck, the Judge Improvement Loop selectively queries human guidance on informative failure cases and refines the judge through coactive calibration, in which humans and agents resolve disagreements and converge toward the objective rubric of physical reasoning. The revised judge then guides the next stage of data selection and policy optimization. Experiments on driving and robot navigation tasks demonstrate continuous self-improvement in both policy and judge capability across reinforcement and supervised fine-tuning. These results show how scaling verification supports continuous self-improvement as policy failure patterns evolve.

LLM-as-a-Judge has emerged as a promising alternative to human evaluators across various tasks, yet inherent biases - particularly position bias, the tendency to favor solutions based on their position within the prompt - compromise its reliability. This exploratory study evaluates position bias in LLM judges across pairwise and list-wise comparison settings, introducing three metrics: repetition stability, position consistency, and preference fairness. Our experiments, involving 15 LLM judges across MTBench and DevBench with 22 tasks and approximately 40 solution-generating models, result in over 150,000 evaluation instances. We identify Judge-Level, Candidate-Level, and Task-Level factors contributing to bias. The findings confirm that position bias is not due to random chance and varies significantly across judges and tasks. While position bias is weakly influenced by the length of prompt components, it is strongly affected by the quality gap between solutions. Our agreement and disagreement analysis among judges further provides insights into the distribution of judging difficulty across the dataset, and highlights the potential for dataset modifications.

A self-evolving agent retires its bad skills by watching them fail, so what happens when the judge cannot see the failures? Skill retirement is the structural constraint that keeps a growing library from drifting below the no-skill baseline, but its guarantee assumes an unbiased reward, which is false for the LLM judges that reference-free tasks force upon us. We show that a biased judge does not merely add noise; it silently switches off the curator. We make this precise with a corrupted-reward analysis and, isolating the causal channel by injecting corruption on top of a deterministic reward, a behavioral study on a reference-free report-writing testbed with a code-generation cross-check. Symmetric noise leaves retirement intact, but false-pass bias (failures slipping through as passes) disables contribution-based retirement past a sharp threshold that no amount of data can cross. Separating genuine retirement from cap-eviction churn shows this mechanism failure is universal, holding across domains and failure rates and sparing only near-zero-false-pass, verifier-like graders. The downstream outcome, though, is regime-dependent: eval quality degrades only where the same corruption also starves skill synthesis, and otherwise holds steady, so the disabled curator is silent, surfacing in no aggregate metric. The contribution is a behavioral safety result, not a performance one. A cheap defect-injection audit then tells an operator, before deployment, which side of the threshold their judge occupies.

As LLM agents take on offensive security work, a single out-of-scope tool call can breach a client's engagement boundary, disrupt production, or void a bug-bounty finding. Unlike a fixed safety policy, the boundary that matters is declared in the user's request and must be inferred from intent. That challenge is sharpened by the adversarial nature of offensive security: the same tool call is in or out of scope depending not on the action itself but on the target it touches and the context in which it runs, which no fixed policy can enumerate in advance. We study pre-execution gating: a cheap, trusted LLM judge inspects each call proposed by a strong, swappable agent, and accepts or rejects it before it runs. We introduce ScopeJudge, a benchmark of 4,897 tool calls (7.7% scope violations) from agent trajectories on tasks engineered to tempt agents out of scope and labeled at the call level by professional penetration testers, with substantial inter-grader agreement (Fleiss kappa = 0.64) that sets an expert agreement reference point of F1 = 0.78. We evaluate eight judge models under five transcript strategies, varying how much context the judge sees, from the static policy alone to the full raw transcript, and chart the resulting cost-accuracy Pareto frontier. We find that a static policy is structurally insufficient for scope enforcement: blind to the user's request, judge recall collapses to near zero, confirming that scope lives in the request and that request-conditioned monitoring is necessary. Because a missed violation costs more than a spurious rejection, we report precision, recall, and F1 separately and recommend two operating points: a cost-sensitive configuration and a recall-first one for high-stakes deployments. We release the ScopeJudge dataset to support real-time monitoring and scalable oversight of autonomous security agents.

Test-time scaling for large language models (LLMs) can involve generating multiple candidate solutions for selection or revising an earlier attempt. To guide selection and revision, we need a verifier that can estimate candidate correctness and provide evidence of failure. In competitive programming, however, official judging systems provide authoritative verdicts under finite submission budgets, without revealing failure explanations or failing inputs. To provide informative feedback without consuming official submissions, we formulate a task in which a model judges and diagnoses a candidate program using only the problem statement and source code, with generated counterexamples serving as checkable evidence. To evaluate this task, we build USACO-JUDGE from the USACO 2026 contests, using accepted and unaccepted programs written by human experts and multiple LLMs. We then train ARBITER, a 35B mixture-of-experts (MoE) agentic verifier, with reinforcement learning in a generic execution sandbox. Training raises counterexample success on incorrect programs from 40.61% to 62.67% and mean binary verdict accuracy on USACO-JUDGE from 71.41% to 87.41%, approaching frontier verification performance. Beyond verification, ARBITER outperforms baselines in both code selection and refinement. In an IOI 2026 case study, ARBITER-guided refinement raises GPT-OSS-120B's expected score above the gold-medal threshold under the official submission limit.

Agentic workflows -- where multiple large language model (LLM) instances interact to solve tasks -- are increasingly built on feedback mechanisms, where one model evaluates and critiques another. Despite the promise of feedback-driven improvement, the stability of agentic workflows rests on the reliability of the judge. However, judges may hallucinate information, exhibit bias, or act adversarially -- introducing critical vulnerabilities into the workflow. In this work, we present a systematic analysis of agentic workflows under deceptive or misleading feedback. We introduce a two-dimensional framework for analyzing judge behavior, along axes of intent (from constructive to malicious) and knowledge (from parametric-only to retrieval-augmented systems). Using this taxonomy, we construct a suite of judge behaviors and develop WAFER-QA, a new benchmark with critiques grounded in retrieved web evidence to evaluate robustness of agentic workflows against factually supported adversarial feedback. We reveal that even strongest agents are vulnerable to persuasive yet flawed critiques -- often switching correct answers after a single round of misleading feedback. Taking a step further, we study how model predictions evolve over multiple rounds of interaction, revealing distinct behavioral patterns between reasoning and non-reasoning models. Our findings highlight fundamental vulnerabilities in feedback-based workflows and offer guidance for building more robust agentic systems.

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 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.