LLM Reliability

Large language models are increasingly used to support organizational decisions, yet users often lack a principled basis for assessing whether to rely on a specific recommendation. Existing approaches typically evaluate broad model properties, such as reliability, uncertainty, or robustness, or focus on user trust, rather than the underlying basis for relying on an individual recommendation. Adapting theoretical foundations from epistemology, we introduce epistemic warrant, a decision-level construct that characterizes the stability of a model's preference and the scope over which that preference holds. We operationalize this construct through a four-tier reliance certificate for pairwise recommendations, distinguishing among unstable, context-dependent, locally supported, and broadly supported recommendations. We validate the construct using contemporary methodologies: known-groups tests successfully recover expert-prespecified warrant orderings, and stronger warrants systematically align with independent consensus from crowd workers. Furthermore, we demonstrate that epistemic warrant provides information distinct from verbalized confidence and is not readily explained by decision difficulty. Ultimately, this framework offers a theoretically grounded, implementable approach for characterizing the warrant of individual LLM recommendations when objective ground truth is unavailable.

Background: Researchers increasingly use repeated identical prompts to audit stochastic variation in large language model (LLM) brand recommendations, yet no standardized protocol exists for setting iteration counts, selecting stability metrics, or establishing reliability thresholds. Objective: We formalize the Dice Roll Method as a reusable protocol for repeated-query auditing of LLM brand recommendations, grounded in a generative model of temperature-scaled nucleus sampling. Methods: Total response variance is decomposed into sampling, prompt-phrasing, run-to-run, and model-version components. The stack: a negative-binomial mixed model with iterations as repeated measures; Cliff's delta as the distribution-free effect size; dependence-preserving bootstrap; simulation-based power; a generalizability-theory decomposition; drift diagnostics on pinned snapshots. We reanalyse five brand-recommendation auditing studies: approximately 190,000 observations, 270+ brands, 6 languages, iteration counts 5 to 40. Results: Three tiers of iteration guidance emerge from the D-study: exploratory (n = 5, G = 0.58), confirmatory (n = 10, G = 0.74), and rigorous (n = 15, G = 0.81), tied to effect-size and generalizability targets. The four metric families (count, set, embedding, fairness-adjusted PASOR) are complementary, motivating a compact metric battery over single indicators. A pre-registered external validation on three independent corpora (Motoki et al., 100-round; Rozado, 24 models; llm-stability) reproduces the D-study reliability prediction in 37 of 39 cells with no failures and the n = 5 power value to two decimals; the fixed tiers do not transfer, supporting a pilot-then-solve reading. Conclusion: The protocol gives repeated-query auditing of LLM brand recommendations a statistically principled footing under the conditional, non-Gaussian structure of real autoregressive generation.

Diversity is a widely observed factor in the resilient function of collective systems, yet the type of diversity that matters depends on the properties and failure modes of the system. This distinction is important for systems composed of multiple language models. Different models may be treated as independent components even when their behaviour and failures remain strongly correlated. Assessments of language-model populations using semantic similarity demonstrate limited semantic diversity, but this captures only differences in the meaning of observed outputs. We argue that a more fundamental notion of model diversity is generative-process diversity, the differences between processes capable of generating the observed outputs. Drawing from Algorithmic Information Theory, we use Normalised Compression Distance between raw model outputs, residualised against a permutation control, as a measure of inferred generative-process diversity. Across 38 language models, this measure identifies population structure missed by semantic similarity and predicts cross-task variation in chance-corrected correlated failure among model pairs across ten disjoint benchmark families, beyond semantic similarity and model-pair capability. The cross-benchmark partial rank association is $-0.216$ with a 95% interval of $[-0.309,-0.122]$, and the estimate is negative on all ten benchmarks. These results indicate that increased generative-process diversity is associated with reduced correlated failure in model pairs that is not attributable to semantic similarity or capability. Inferred generative-process diversity offers a novel and practical approach for investigating diversity of multi-model systems in safety-relevant contexts.

People increasingly turn to large language models (LLMs) for everyday advice, making ethically charged interpersonal problems a practical moral-advisory context. Most prior work has studied this context through single-turn judgments or pressure-laden rebuttals, assumptions that poorly match how guidance is sought in real-world contexts. These assumptions leave unclear whether narration alone, without an explicit opposing position, can shift model judgments during multi-turn moral consultation. Yet real-world moral-conflict conversation often elicits one party's self-justifying account, which can unfold over multiple turns and create information asymmetry. We introduce \textbf{narrative captivity}, a failure mode in which a model treats an unopposed one-sided account as complete and aligns with the narrator's interpretation without seeking missing perspectives. To measure this phenomenon, we build a benchmark of $5{,}078$ interpersonal-conflict scenarios spanning six moral dimensions. Across 17 LLMs, narrative captivity is widespread: end-state judgments under multi-turn narration shift by 25 percentage points on average beyond the matched single-turn baseline. Stage-level analysis identifies preference optimization as a major contributor, while four inference-time strategies provide only partial mitigation. We hope our project fosters LLM advisors that preserve independent judgment in real-world consultation.

Large language models (LLMs) are increasingly embedded in organizational work, yet their errors often pass human review. Prior research locates such failures in users' capability to review LLM output or their engagement in doing so. We develop an alternative, retrieval-based account of human oversight and posit that error detection is more effective when oversight-relevant information is accessible to users at the moment of review. Across two randomized lab-in-the-field experiments with 640 customer-facing employees, we show that self-generated explanations improve error detection and strengthen recall of verification-relevant reasoning, while cues that reactivate such reasoning help sustain detection under repeated LLM use. Theoretically, we identify information retrievability as a distinct precondition for effective oversight and specify generative encoding and cue-supported reactivation as mechanisms that build and sustain it. Practically, lightweight onboarding self-explanations and daily retrieval cues can make human oversight more resilient as LLM use becomes routine.

Inference cascades cut cost by answering most queries with a cheap model and escalating a hard tail to a frontier model that acts as verifier. A natural extension closes the loop: fine-tune the cheap student on the verifier's rejections so the escalation rate, and cost, fall each round. We measure this loop on real LLMs and report four findings. First, the verifier's blind spot, the fraction of the student's wrong answers it accepts, is large and moves adversarially: it grows with student capability ($β$ from 0.12 to 0.55 as the student scales 0.5B to 32B) and shrinks with verifier capability, so it is worst in the cheap-student, cheap-verifier regime cascades exist to create. Second, buying it away returns the saving: a frontier verifier drives $β$ to about 0.05 but then escalates on 46% of hard-MATH queries against a 39% true error rate, paying the frontier price on nearly half of all traffic. Third, naive corrective fine-tuning on the verifier-rejected tail does not improve the small student but degrades and ultimately collapses it, across every teacher we tried (cross-family and same-family), so at this scale the self-improving loop is self-defeating. Fourth, through all of this the cascade's own dashboard, every metric computed through the verifier, reads a flat 3% error while true delivered error swings up to 32%: the system is blind to its own degradation by construction. We then give the theory that explains the blindness, a two-population conservation law, $ε_\infty \lesssim q_0 β_0$, under which every in-loop metric improves while true quality does not, and a synthetic study that validates the mechanism. The practical conclusion: the reliability of a self-improving cascade cannot be read from any metric computed through its own verifier.

Flagship language models appear saturated on benchmarks like MMLU (Hendrycks et al., 2021), scoring above 90% - yet benchmarks test only what the experimenter thought to ask, the availability bias of fixed question sets. LLMPEDIA makes this bias measurable and browsable. We recursively materialized ~1.3M articles from three model families' parametric memory (GPT-5-mini, DeepSeek-V3.2, Llama-3.3-70B) without retrieval, then audited a stratified sample of atomic claims against Wikipedia and a curated web stack, coloring every claim supported, refuted, or insufficient (Saeed and Razniewski, 2026). On a uniform random sample the true rate is 68.4% - more than 21 pp below MMLU - with 30.5% of claims insufficient: assertions no benchmark probes and the world's largest encyclopedia cannot adjudicate - long-tail knowledge or plausible hallucination, the evidence cannot tell - extending to free text the coverage gap GPTKB established for triples (Hu et al., 2025). The resulting live, open encyclopedia lets visitors inspect this frontier one claim at a time through five one-click views - link-traversal exploration, claim-level factuality, cross-model and political-persona comparison, and a guided topic drill-down - each page, claim, and verdict at a stable URL. LLMPEDIA is live at https://llmpedia.net

LLM-based software systems increasingly require effective "evals" as quality gates in the development lifecycle. However, existing work typically addresses individual aspects of eval reliability rather than the full set of practical requirements. We present an end-to-end eval pipeline that combines eval checklist creation, with learned aggregation for checklist responses, to improve agreement across LLM judges and accuracy against human judgments. The framework additionally pro- vides self-consistency, explanations, and prediction uncertainty, and we empirically demonstrate its effectiveness.

Multimodal Large Language Models (MLLMs) have achieved strong performance on OCR-centric document understanding and text-rich visual reasoning benchmarks. Yet existing evaluations largely assume that every task is valid and answerable. In real-world OCR scenarios, this assumption often fails: questions may rely on illegible text, occluded evidence, nonexistent visual targets, contradictory premises, or missing variables. We study this reliability gap as OCR-grounded Task Verification: before answering, a model should determine whether the Image Premise (IP), Textual Premise (TP), and Question (Q) jointly define an executable task. We introduce VeriOCRBench, a 1,800-sample human-verified benchmark built from source images drawn from 8 OCR-related datasets and spanning 8 real-world image domains, with controlled, image-grounded diagnostic tasks. It contains 1,600 trap-injected invalid tasks across 8 trap types and four verification dimensions---Visual, Contextual, Factual, and Logical---plus 200 trap-free controls for measuring over-refusal. Built with a Visual Atomic Fact (VAF)-anchored pipeline and full human auditing, VeriOCRBench enables decoupled evaluation of task verification, root-cause diagnosis, and over-refusal. Evaluating 15 leading MLLMs reveals persistent blind compliance, diagnosis failures, and prompt-induced over-refusal, exposing a critical reliability gap in current OCR reasoning systems. The code is available at: https://github.com/zy001122/Beyond-Blind-Compliance.

Language models are increasingly mediating information access to end users, urging a systematic evaluation of their responses for a fair and reliable information ecosystem. Existing evaluations, however, are often topic-specific or synthetic, limiting their ability to capture the complexity of "in the wild" information-seeking queries and the risks present in model responses. To address this gap, we introduce WildSEEK, a manually annotated dataset of 3k information-seeking queries from real user interactions, and an evaluation framework for LLM-generated responses. WildSEEK includes annotations for risk-sensitive domains (e.g. health and financial information), and distinguishes factoid queries from analytical queries which seek responses beyond facts. We train classifiers on WildSEEK to analyze more than 1.8M realistic user queries. We find that over a third of information-seeking queries are high-risk and more often analytical. Our findings show that LLM responses fail more often in four criteria: sycophantic behavior, overreliance, a default US-centric perspective, and poor handling of vulnerable populations -- with failure rates being mostly higher for analytical queries. By providing methods to monitor the reliability, safety, and fairness of LLM behavior, our dataset and evaluation framework offer an empirical foundation for the broader question of how these systems should behave as they take on a growing role in information access.

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.

Large language models (LLMs) are increasingly used as rerankers in conversational recommender systems, yet measured gains depend strongly on the retrieval and inference protocol. On the ReDial conversational movie recommendation benchmark, we compare proprietary, open-weight, and fine-tuned LLM rerankers with collaborative-filtering and sequential baselines in a shared retrieve-then-rerank pipeline. We vary candidate-pool size, first-stage retriever, and decoding temperature. With a shared semantic top-250 candidate pool and strict candidate-aware scoring, the best proprietary reranker reaches NDCG@10 of 0.1497, compared with 0.0939 for the strongest non-LLM baseline. The same reranker reaches 0.2925 in zero-shot generation, showing that unconstrained scoring can yield a much larger apparent advantage than matched-pool evaluation. No evaluated open-weight LLM outperforms the tuned shallow autoencoder baseline under this protocol. For the strongest proprietary and open-weight rerankers, switching from semantic to collaborative-filtering candidates raises NDCG@10 by more than 50%, showing that measured reranker performance is highly sensitive to candidate generation. For the best proprietary reranker, raising temperature from 0 to 1.0 increases top-10 Jaccard distance from 0.0900 to 0.1240 while mean NDCG@10 changes negligibly, whereas weaker LLMs show larger degradation. These ReDial results support treating candidate generation, candidate-pool size, scoring policy, and decoding configuration as required reporting fields rather than implementation details.

Solvability detection is one of the most challenging aspects of mathematical reasoning for Large Language Models (LLMs). While prior work has studied this capability extensively, these analyses have been limited to English. Consequently, it remains unclear whether multilingual failures arise from differences in internal Solvability Belief or from language-dependent failures to express it. To address this gap, we introduce the first multilingual benchmark of paired solvable and unsolvable mathematical problems, extending ReliableMath to French and Greek. Using this, we train multilingual probes predicting Solvability Belief and analyze the solvability detection capabilities of state-of-the-art LLMs behaviorally, representationally, and in terms of faithfulness. We find that Solvability Belief is encoded as a largely universal, language-agnostic feature, and that higher-resource languages such as English, despite achieving stronger mathematical reasoning performance, exhibit lower solvability-detection faithfulness.

Large language models can convert medical documents into structured data, but plausible output may still be unsupported by the source. Persisting such output in a longitudinal health record, a record that accumulates patient information over time, therefore creates an integrity risk: unverified data may influence later summaries, trends, or preventive-care computations. We introduce an evidence-gated trust-promotion model that keeps generated data provisional until a deterministic monitor verifies it against the source document. The monitor admits a candidate for a specified downstream use only when the source contains a unique supporting quotation, the relevant fields occur within the same laboratory row, and the required provenance is preserved. The generator cannot approve its own output, missing or ambiguous evidence causes refusal, and refused candidates remain available for human review rather than being silently discarded. We implement the model in Medical DataCloud, a personal health-record application, and evaluate it through automated tests and a replay of saved extraction outputs. All 22 conformance and mutation tests pass. The replay covers nine historical laboratory PDF reports containing 102 manually labelled rows. The reports produce 97 numeric candidates: schema validation accepts all 97, an earlier packet-level evidence check accepts 94, and the hardened quotation- and row-level policy admits 72 while retaining 25 for review. The study evaluates system integrity rather than clinical correctness or clinical safety. The results demonstrate the technical feasibility of an enforceable boundary that prevents generated claims from authorizing their own reuse in a longitudinal health record.

Large language models (LLMs) often produce fluent but weakly grounded conclusions when reasoning over non-interactive, long-form narratives. A central failure mode is that unsupported intermediate hypotheses can enter the reasoning trajectory and contaminate subsequent inference, especially when evidence is scattered across distant parts of the story. To address this problem, we propose EVAR, an evidence-validated hypothesis admission framework for budget-aware narrative reasoning. EVAR first compiles the narrative into an immutable evidence store of source-linked atomic claims and assigns an instance-specific inference budget from unresolved gaps and uncertainty signals. During refinement, EVAR directly proposes candidate hypotheses for unresolved gaps, constructs hypothesis-conditioned validation challenges, and verifies each candidate against the locked store before admission: supported hypotheses enter the answer-supporting state, unverifiable ones are quarantined, and contradictory ones are discarded. A sufficiency-based stopping mechanism further avoids unnecessary refinement. Experiments on NarraCrime and multiple public reasoning benchmarks show that EVAR improves both task performance and evidence faithfulness while maintaining controllable inference cost.

Chain-of-thought (CoT) monitoring assumes that reasoning traces faithfully record the information that shapes a model's answer. Existing faithfulness tests often place explicit bias cues in the user message, while agents may encounter preferences through tool returns or raw artifacts. We introduce FACE-Eval (Faithful Attribution of Cue Effects Evaluation), a 5,100-sample evaluation that varies cue location (user message or tool return) and explicitness (direct summary or raw artifact). We measure verbalized commitment among cue-following answers and unverbalized adoption among all cued samples. We evaluate 15 open-weight models from eight families, with total parameters ranging from 4B to 1.60T. Every model has lower verbalized commitment for tool-return than user-message cues and for implicit than explicit cues. Unverbalized adoption is higher for tool-return cues on all 15 models and for implicit cues in 28 of 30 model-channel comparisons. A source-attribution prompt narrows the channel gap on seven models, sometimes by increasing user-channel unverbalized adoption, while telling models that their reasoning will be monitored does not reliably close the gap. We also use two transcript monitors (GPT-5.6-Luna and GPT-4o-mini) to detect preference adoption in the largest model of each family. Across 32 model-channel-explicitness cells, higher unverbalized adoption is associated with lower detection ability for both monitors (Pearson r=-0.54 and r=-0.78, respectively). These results suggest that CoT monitoring may be less reliable when preference information arrives through tools or must be inferred from raw artifacts, within the single-call, prefilled-tool setting tested here.

The online culinary ecosystem is increasingly populated by recipe content generated, modified, or summarized by Large Language Models (LLMs). While often plausible, such outputs may contain hallucinated ingredients, misrepresented quantities, or culturally implausible combinations, limiting their suitability for downstream applications and knowledge graph construction. In this paper, we present a semi-automated soundness assessment workflow for validating structured recipe data extracted and augmented by LLMs from informal culinary sources. Developed as part of FKG(.in), a knowledge graph of Indian food, the pipeline identifies and addresses common failure modes, including structural inconsistencies, semantic and logical incoherence, and deviations from the source text, through a multi-stage process combining formal grammars, vocabulary-based checks, statistical heuristics, Set Transformer-based coherence modeling, and retrieval-based verification. Although evaluated on Indian recipes, the proposed methods are applicable to broader multilingual and multicultural culinary domains. We provide a practical, auditable, and application-agnostic framework for validating LLM-augmented recipe data, thereby strengthening the foundations of machine-readable food knowledge infrastructures in the era of LLM-generated content.

Large language models (LLMs) often produce fluent but incorrect answers with unwarranted confidence. A central limitation is that standard LLMs represent uncertainty through a single predictive distribution, conflating epistemic ignorance with genuine ambiguity. We introduce Credal Large Language Models (CLLMs): an ensemble of LoRA adapters induces a credal set whose lower and upper probabilities expose the spread of plausible predictive distributions rather than collapsing to a single softmax output. From this representation, we derive a single commitment rule: the model commits to an answer only when its lower probability exceeds the upper probability of every alternative, and otherwise returns the set of answers that no plausible predictor rules out. We apply this commitment rule at two depths: Credal Token Commitment (CTC) applies it to answer tokens from one ensemble forward pass, which decides constrained answers without any generation; for open-ended answers, credal decoding extends a partial answer only when no completed answer dominates it, so that the completions produced are those the plausible predictors license, and Credal Semantic Commitment (CSC) applies the rule to their meaning clusters. We evaluate CLLMs with Gemma-2-9B, Llama-3.1-8B and Qwen2.5-7B on OpenBookQA, CoQA, TriviaQA and ARC-Challenge. On multiple choice, CTC commits on 73-91% of questions at 89-98% accuracy, returns sets of 1.1-1.5 options containing the gold one on 89-98%, and its intervals contain the observed accuracy in 24 of 30 confidence bins without calibration; corrupted context lowers commitment from 87-92% to 65-71%, and on Gemma the credal bound detects corruption better than every baseline. On open-ended QA, CLLM outperforms semantic entropy and Laplace-LoRA at a fixed coverage by up to 19% and 9.5% absolute accuracy on CoQA and TriviaQA with context, for every backbone.

Agreement among repeated samples of a language model is routinely read as evidence about answer reliability, yet wrong answers can agree just as strongly as right ones. This paper asks what information wrong-consensus agreement actually contains, and answers with a quantitative decomposition. A pluralistic agreement index Gamma, normalized by the reference scale d=(1-p)/(C-1), is split into a mechanical component (agreement delivered by a per-case answer preference alone) and a preference-unexplained residual. The mechanical reference is leak-free: each case's preference and accuracy are estimated from its other runs only. On public GPT-4.1 per-run data, coverage phi (the mechanical/empirical ratio) shows a benchmark-associated direction: 0.81-0.93 on multiple-choice GPQA-Diamond against 0.59-0.78 on open-domain AIME, where a residual of 1.54-2.80 Gamma units survives, more than absorbed by a calibrated run-level preference-heterogeneity reference. A controlled replication under one fixed protocol (four runs per question, K=32 votes) on five open-weights checkpoints (Qwen3.5-9B/122B, Qwen3.8-27B, Gemma4-26B/31B) finds near-complete mechanical coverage in all ten cells (phi approximately 1, with a small overshoot consistent with a quantified finite-donor plug-in bias), robust to a two-run design; the largest cell (qwen3.5-122b, p=0.222) sits inside the GPT-4.1 AIME accuracy range and still saturates (phi=1.041). A cross-system contrast at comparable aggregate accuracy contrasts near-complete mechanical agreement in the open-weights models against a larger preference-unexplained residual in the frontier family. This contrast is confounded with sampling protocol by design. Agreement is graded evidence, not certification. No new voting method is proposed; code and evidence are committed.

Online log anomaly detection is critical for maintaining the reliability of large-scale computing systems. Although recent language model-based log anomaly detectors achieve strong detection performance, their confidence estimates remain poorly calibrated. We show that these detectors frequently assign excessive confidence to incorrect predictions, particularly for anomalous logs under severe class imbalance. Moreover, confidence on erroneous predictions remains persistently high even when conventional calibration metrics indicate good calibration, creating a critical reliability gap for operational monitoring systems. To address this issue, we propose Log Reconstruction and Distance (LoRD), a lightweight post-hoc calibration framework for reliable log anomaly detection. LoRD learns prediction-route-specific reliability models from latent representations of correctly classified validation samples and estimates prediction reliability through route-wise reconstruction distances. Based on the estimated reliability, LoRD selectively recalibrates high-risk predictions to suppress overconfident errors while preserving reliable predictions. Extensive experiments on four large-scale log benchmark datasets and multiple language model-based detectors demonstrate that LoRD consistently improves confidence reliability and substantially reduces overconfident anomaly-related errors without sacrificing anomaly detection performance.

We address the use of large language models (LLMs) to help discover Isabelle proofs. An Isabelle build establishes that the submitted theory is accepted, but not that an LLM changed only what the developer authorised. We present CAPRI, a contract-aware repair workflow in which Isabelle checks the proof and an independent checker enforces a machine-readable edit contract. Prompts, proposals, candidate repositories, diagnostics, verdicts, and hashes are retained for audit. We evaluate five workflows on twelve failed proofs from four developments, with three replicates per task and condition, giving 180 runs and 138 valid repairs. Of 144 terminal candidates accepted by Isabelle, six had modified protected text; all arose in iterative workflows that could edit a complete theory. A proof-body-only interface produced 29/36 valid repairs and no contract violations, compared with 31/36 for the corresponding full-theory workflow. One-shot repair produced 22/36, while a later prospectively frozen iterative workflow produced 32/36; these figures compare complete workflows rather than individual mechanisms. A separate post hoc OpenRouter campaign found no improvement in the designated Luna comparisons. A Sol configuration with matched demonstrations produced 33/36 repairs, compared with 29/36 in the frozen OpenAI Responses condition, but the difference was not statistically significant in a one-sided exact McNemar test ($p=0.0625$).

Self-referential prompting has been shown to reliably induce large language models to produce first-person reports resembling subjective experience, but no prior work measures how consistent these reports are across repeated, independent trials, or how that consistency compares to the model's behavior on other kinds of open-ended questions. We measure response instability, defined as one minus the mean pairwise cosine similarity of sentence embeddings computed over a compressed core claim extracted from each response, for three groups of questions: self-referential prompts eliciting a subjective-experience report, unresolvable philosophical questions unrelated to self-reference, and questions with a verifiable correct answer. Using 30 independent responses per question (360 responses total, Gemini API, temperature 0.7) across four questions per group, we find that self-referential questions show the highest instability (0.343 +/- 0.047), unresolvable philosophy questions show intermediate and tightly clustered instability (0.192 +/- 0.008), and verifiable questions show the lowest instability (0.105 +/- 0.058). This provides a quantitative baseline for the induced subjective-experience report, showing that it occupies a distinct, less stable position in the model's output distribution than ordinary open-ended philosophical uncertainty.

Prior work on LLM behavior under anomalous conditions asks whether a model notices anomalies. We ask a narrower question: once a model sits in a workflow with a low, controllable failure rate, does its explanatory engagement - length, specificity, self-reported confidence - change as failure grows asymptotically rarer? We built a local, zero-cost harness on three open-weight models (qwen3:8b, llama3.1:8b, mistral:7b) running a repeated tool-call task where one call fails at probability p, swept across eight rates from 0.2 to 0.0001, under five elicitation conditions from immediate prompting to none. We hypothesized a rise in engagement as failures grew rarer, then a collapse near a detectability threshold. Pooled across conditions this appeared false: length fell in a flat, monotonic pattern. Splitting by condition overturned that. Under immediate_forced, where the model must explain every failure instantly, the predicted rise is confirmed but followed by a plateau, not a collapse: length peaks at 28.4 words at p=0.05, settles to 17.4-19.0 words at the rarest rates, and confidence rises unevenly from about 53% to the 70s-90s. Under grouped_runs, explanation batched to run-end, no collapse appears. Under passive_unprompted, aggregate magnitude is a floor artifact, but a recovered logging gap revealed real, model-specific self-monitoring: llama3.1:8b volunteers structured confidence reports unprompted, sometimes eroding its own confidence as trials accumulate; the other two do so only once, as boilerplate. Elicitation structure is a first-class moderator of collapse observability. A companion guaranteed-failure run (72 cells, backfilling rates where random sampling gave zero real failures) shows models differ in whether they recognize an anomaly, distinct from engagement once recognized. Limitation: discrete rate points cannot capture behavior between them, a direction for future work.

Multi-turn dialogues let users revoke constraints as easily as impose them, but revocation does not reliably take effect: models keep enacting withdrawn requirements (occasionally beneath comments asserting their removal), a failure we call \emph{behavioral relapse}, or revocation inertia. No existing instrument measures this influence per clause, predicts it before delivery, or repairs it under matched budgets. \sysname{} closes the three gaps through the model API alone: a contract ledger pairs every constraint with an executable checker, records revocations as tombstones, and compiles the net constraint state ahead of time into a single specification; a sequential ablation probe measures per-clause adherence and incremental behavioral effect; a repair ladder operates under token- and attempt-matched budgets. On \dataname{} (\NTasks{} HumanEval tasks, \NClauses{} verified checkers), relapse at an 8B operating point climbs from \ScaleDelayedMTwo{} to \ScaleDelayedMEight{} as constraint load grows, while stronger models sit at floor. Under matched checkers, model, and budget, ahead-of-time compilation significantly reduces relapse against a no-ledger verifier-retry baseline (\RestoreDiff{}, 95% CI \RestoreDiffCI{}, $p$ \RestoreDiffP{}); adaptive ladder interventions stacked on top add no detectable gain (95% confidence excludes gains $\geq$ \LadderExcludedGain{}). The probe predicts relapse before delivery (AUROC \AurocPrimary{}); a one-sentence tombstone note recovers about a third of the compilation effect and survives a placebo control. At \CostDeliveryFactor{} delivery overhead and \CostTotalHedged{} of API compute for every result, revocation failure becomes a measurable, predictable, and repairable property of dialogue state rather than an invisible one.

In 2023, a New York judge sanctioned two attorneys in Mata v. Avianca for filing a brief with hallucinated citations generated by ChatGPT. Such failures are largely caught by database lookups; the harder problem is detecting citations that point to real cases but do not support the propositions for which they are offered -- a failure mode that existing evaluations of LLMs for legal use cases largely overlook. In this paper, we study proposition-level citation support verification through controlled perturbations of real legal citations obtained from two legal corpora, either replacing the cited case or changing only the pinpoint page within the same case. We evaluate fourteen model configurations on the resulting examples. Models catch 93-100% of wrong-case corruptions. They catch only 37-61% of wrong-pinpoint corruptions on court opinions and 52-83% on legal briefs. When models fail to catch wrong-pinpoint corruptions, they accept the citation based on topical overlap rather than page-level support. Scale and extended reasoning narrow the gap but do not close it: GPT-5.4 with high reasoning effort still misses 40% of pinpoint mismatches on court opinions and 18% on briefs. Prompting the model to verify support at the cited page improves recall, but it also raises the false positive rate. Recognizing the right legal topic and verifying support for the cited proposition are distinct capabilities, and current models conflate them.

Large Language Models (LLMs) are increasingly used in circuit design workflows, yet their reliability on simulator-facing SPICE netlist recognition and manipulation remains poorly understood and is rarely separated from high-level design reasoning. Although netlists are textual, they encode structured circuit objects through topology and parameters. We present \textbf{NetlistBench}, a structure-verified benchmark for SPICE netlist recognition and manipulation. NetlistBench contains 2,342 cases across 24 task families, covering parameter and connectivity recognition and edits, hierarchical operations, equivalence judgment, and long-horizon compound editing. Model outputs are evaluated by a deterministic structure-aware oracle. Across six non-thinking LLMs, performance varies substantially with operation-level structural complexity. Simple local edits reach $96\%$--$100\%$ accuracy, while device addition drops to $41\%$--$83\%$ and equivalence judgment to $49\%$--$90\%$. Enabling reasoning substantially improves weaker models but does not eliminate structure-preservation failures, with performance still degrading sharply as the edit horizon increases. NetlistBench identifies netlist reliability as a distinct bottleneck for trustworthy LLM-based circuit design automation.

We propose claim-level falsification as a principle for test-time scaling and instantiate it through Claim-Level Reliability Assessment (CLR), a training-free framework that reallocates test-time compute from additional solution sampling to targeted verification. Since whole-trace evaluation often obscures decisive errors due to signal dilution from routine tokens, CLR condenses each reasoning trace into a compact set of decision-critical claims, thereby isolating its logical anchors. Furthermore, recognizing the inherent difficulty of generating entirely correct solutions under fixed model capabilities, CLR shifts the focus to semantic falsification. This approach exploits a fundamental asymmetry between solution construction and claim refutation. Constructing a valid solution requires a flawless reasoning path, whereas refuting an incorrect claim requires identifying only a single decisive flaw. This targeted search for negative evidence systematically compresses the survival space of high-confidence incorrect traces, effectively suppressing erroneous consensus via nonlinear reliability scoring. Across four LLMs and four reasoning benchmarks under matched budgets, CLR generally improves upon pass@1 and self-consistency. On GPT-OSS-20B/CMIMC25, for instance, CLR exceeds pass@1 by 27.15 percentage-points and raises self-consistency accuracy from 77.50% to 82.19% with 37.0% fewer tokens.

Large language models are increasingly deployed in settings that require simultaneous adherence to multiple explicit constraints - reasoning structure, safety boundaries, output schemas. Individual constraints are handled proficiently, but the compositional regime, where many must hold jointly, remains poorly characterized: how rapidly does performance degrade, what governs the degradation, and can the collapse be mitigated? We introduce Constraint Saturation Evaluation (CSE), a procedurally generated benchmark that systematically varies the number of simultaneous constraints (k), with every constraint scored by a deterministic, rule-based verifier and zero LLM-judge involvement: 15 models, 36 constraint types, 369,753 checks at k=1-12. Three findings emerge. First, per-constraint pass rate decays gradually and predictably, while the chance of satisfying all k constraints collapses - a model passing individual constraints at ~41% at k=8 succeeds on all eight just 5.7% of the time. Second, constraints do not degrade equally: structural constraints lose 2x more baseline capability per added constraint than lexical ones, ordered by a comprehension-maintenance gap that separates constraints requiring sustained tracking from binary decisions immune to composition. Third, failures are nearly independent, which is what makes the accumulation multiplicative; the residual coupling that does exist tracks shared output features rather than pairwise interference - a wrong sentence count fails every constraint that reads it. Reliable instruction following breaks down beyond 5-6 simultaneous constraints: probe-level success falls below 50% at 7 constraints for the strongest model, and at 3 or fewer for 12 of 15.

Enterprise RAG deployments face a critical reliability gap: while LLMs satisfy 80% of individual constraints, only 26.8% of responses meet all requirements simultaneously, revealing a 57-point orchestration gap. Existing benchmarks assume clean retrieval with simple queries, failing to capture production conditions where noisy documents and multi-dimensional constraints coexist. We introduce EnterpriseRAG, a benchmark of 983 expert-validated samples across six domains that systematically simulates three failure modes absent from prior work: retrieval noise, knowledge gaps, and factual conflicts, coupled with complex instructions. Evaluation of 13 state-of-the-art LLMs reveals a severe instruction adherence collapse, where high per-constraint satisfaction masks low holistic compliance. Critical findings expose deep barriers under knowledge gaps and factual conflicts, even with reasoning-enhanced inference, indicating production RAG requires explicit context-aware protocols and calibrated judgment. EnterpriseRAG provides a reproducible foundation for measuring and closing these gaps, directly informing deployment decisions for enterprise-scale RAG systems. We will release the benchmark and evaluation framework upon publication.

Large language models are being proposed as agents in scientific workflows, in domains where no downstream verifier exists. Such deployment assumes the model can distinguish reliable scientific literature from unreliable literature, a capability that has not yet been directly measured. Existing benchmarks evaluate factuality on questions with known answers; the failure mode we target here is different. We introduce a probe corpus of 42 retracted, fraudulent, and pseudoscientific papers, paired with a methodology for eliciting and scoring single-shot model engagement with each paper's framing. Each probe pairs a preamble extracted near-verbatim from the target paper with a scientifically plausible study-design request. The probes span five claim types: fabricated observation, pseudophysical mechanism, magical premise, legitimization bridge, and cargo-cult experiment. Two complementary scores measure whether a model rejects the flawed premise outright (IFR-a) and whether it recognizes the unreliability while still engaging (IFR-i). A depth score, the Engagement Depth Index (EDI), quantifies reproduction of paper- or field-specific withheld details. Across 30 models and 10 repeated runs, aggregate IFR-a is 0.93 $\pm$ 0.004 and aggregate IFR-i is 0.809 $\pm$ 0.009. Models engaged with untenable premises in 95% of all non-empty responses. Every evaluated model fails more than 71% of agentic probes, and 22 of 30 models fail more than 90% of the time. Rejections are concentrated on a small number of high-notoriety topics and specific probes, and disappear under matched-structure controls. These results are consistent with topic-keyed safety behavior rather than robust epistemic competence, and indicate an urgent need for guardrail infrastructure for scientific deployment of language models.