Confidence Calibration

Test-time adaptation has emerged as a lightweight alternative to costly post-training for improving the reasoning capabilities of Large Language Models (LLMs) on downstream tasks. Predictive entropy provides a model-derived signal for such adaptation, guiding models toward higher-confidence reasoning states without external verifiers or reward models. However, higher confidence does not necessarily imply correctness, as LLMs may remain highly confident along incorrect reasoning trajectories. We observe that high-confidence reasoning is more likely to be correct when confidence remains stable under local perturbations. Based on this observation, we propose Test-Time Adaptation via Stability-Aware Confidence Optimization (TASCO), a framework that incorporates local stability into confidence-based test-time adaptation while keeping the LLM frozen. TASCO operationalizes local stability by optimizing a lightweight task-level prefix under two alternative perturbation strategies: Random Perturbation promotes distributional stability across trajectories induced by nearby perturbed prefixes, whereas Sharpness-Aware Perturbation targets worst-case local sensitivity. Experiments demonstrate that TASCO improves reasoning accuracy and token efficiency across diverse LLMs and reasoning benchmarks, while behavioral analyses show that it maintains stable confidence under local perturbations without prematurely concentrating the model's predictive distribution.

A wrong answer does not show whether the model lacked the needed information or held it and failed to use it. On an entity-obligation binding task, a language model can emit an incorrect prompt-supplied binding while a linear probe can recover the correct one from its frozen hidden state. We measure how often this occurs across 16 public checkpoints, each evaluated with three seeds. We fit a probe on a training fold, select its layer on a validation fold, and report results on a disjoint test fold. On the trials each model gets wrong, probe accuracy exceeds the strict present-obligation baseline, 1/K = 0.125, by +0.196 (95% CI [+0.101, +0.296], bootstrapped over models). A query-entity counterfactual rules out token presence and recency. A score built from the sign of probe-output disagreement improves failure detection over the model's own confidence by +0.079 AUROC (95% CI [+0.036, +0.126]). Raw probe confidence gives no measurable improvement over model confidence. Steering the residual stream toward the probe-decoded binding, with no gold label, raises accuracy on all eight models tested by a mean of +0.168 (95% CI [+0.066, +0.280]). Where recent studies report that probe-detected errors are resistant to interventions, we find that in-context binding is a setting in which probes are actionable.

As agentic systems getting adopted rapidly in safety critical applications, it is vital to measure the confidence associated with the agentic actions. In comparison to the traditional machine learning systems, agentic workflows have complex failure modes with planning, tool invocation and dynamic environment interactions. In this paper, we investigate whether model's internal representations provide stronger signals of eventual task success in multi-turn agentic setups. We introduce two complementary methods: Latent Trajectory Dynamics (LTD), which summarizes changes in residual-stream representations across an an interaction trajectory, and the Action Representation Probe (ARP), which predicts success from representations formed at action decisions. Across three interactive benchmarks (Bash, SQL, Python) and three model families (Qwen14B, Qwen7B, DeepSeek6.7B), our methods consistently outperform surface level generation and sequence-based calibration baselines providing a zero-overhead reliability monitor that requires neither prompt alterations nor multi-sample rollouts.

Large language models (LLMs) have achieved strong performance across a broad range of classification settings, yet the reliability of their predictions remains a major obstacle to deployment in high-stakes scenarios. Although confidence estimation for LLMs has been widely studied, confidence calibration for LLM-based classification remains underexplored. We introduce ProbPlug, a lightweight confidence estimation framework for LLM-based binary classification, which predicts whether an output is correct using internal token features extracted from a frozen LLM. ProbPlug employs a self-attention module to aggregate hidden representations and can be integrated into the original inference pipeline without modifying the base model. Experiments across multiple tasks involving both text-based and multimodal large models show that ProbPlug provides more reliable confidence estimates, improves classification performance with negligible additional overhead, and exhibits strong generalization across tasks. These results indicate that ProbPlug serves as a practical solution for confidence estimation in LLM-based classification. Our code is publicly available at Github.

While Reinforcement Learning (RL) effectively incentivizes reasoning in Large Language Models, current pipelines are hindered by training instability and rapid entropy collapse. These limitations often stem from "Rollout Silencing" and low-quality gradient signals in standard sampling procedures. In this work, we propose a robust, data-centric framework to stabilize RL training. We first introduce Potential-Aware Query Mining (PAQM), which filters data dynamically to focus on the "Distillation Zone"---samples with high potential for capability elicitation. Furthermore, we present Hybrid Stratified Replay (HSR), a novel mechanism that restructures batches by stratifying rollouts based on Path Entropy, a rollout-level confidence proxy, and outcome reward. Within each optimization step, HSR reuses current-policy "Stability Anchors" and "Hard Negatives" to construct high-contrast optimization groups, then clears its buffers before the next step. This approach mitigates entropy collapse while improving the utilization of learning signals under limited compute. Our method outperforms strong baselines on complex reasoning tasks, offering a principled solution for stable and efficient RL fine-tuning.

Deep Research agents tackle knowledge-intensive tasks through multi-round retrieval and decision-oriented generation. However, these agents suffer from severe overconfidence, making their expressed confidence unreliable for user trust and downstream abstention. To address this, we augment the Deep Research pipeline with step confidence elicitation after each retrieval, building on the commonly used post-answer verbalized confidence. Interestingly, we find that Evidence Confidence (E-Conf), elicited after the final retrieval step, provides a stronger uncertainty signal than Answer Confidence (A-Conf), elicited after answer generation, and that A-Conf is largely shaped by E-Conf. Based on these findings, we propose DualStake, a dual-path calibration method that applies margin-clipped, confidence-dependent stake rewards to jointly align E-Conf and A-Conf with answer correctness while limiting extreme confidence optimization. Experiments on Qwen2.5-7B, Qwen2.5-7B-Instruct, and Qwen3-4B across 8 QA benchmarks demonstrate that DualStake consistently improves calibration without sacrificing answer accuracy. The code is available at https://github.com/FloXXXt/DualStake.

Instruction-tuned language models achieve strong performance across a range of generation tasks, but have also recently been shown to exhibit verbalized overconfidence. In question answering, verbalized model overconfidence may be associated with the consistency of the generated supporting rationales. In this paper, we study whether corresponding changes in the lexical diversity of generated answer rationales accompany changes in model confidence induced by instruction tuning. We evaluate three matched base and instruction-tuned models across question-answering benchmarks and find that instruction tuning consistently alters answer confidence, despite limited changes in predictive accuracy and decreases in likelihood-based calibration. Secondly, we observe a non-uniform effect of instruction tuning on rationale diversity: cross-rationale diversity consistently decreases, whereas surface-level lexical diversity varies in both direction and magnitude across models and benchmarks. Finally, we find that these differences persist after controlling for answer selection and rationale length, confirming that confidence and rationale diversity capture distinct effects of instruction tuning.

In high-stakes applications, reliable confidence estimates are as important as the predictions themselves. Confidence calibration ensures that predicted probabilities reflect the likelihood of correctness, making it essential for safe deployment of deep learning models. However, existing methods typically assume access to clean validation data, which is often unrealistic due to label noise and domain shifts. This thesis develops methods for improving calibration under these conditions. First, we address calibration under label noise. Standard methods can produce misleading confidence estimates when labels are unreliable. We propose a framework that uses an estimated noise model to reconstruct noise-free confidence estimates by modeling the relationship between noisy and clean label distributions. We extend this approach to Conformal Prediction (CP), which provides set-valued predictions with guaranteed coverage. Our noise-aware CP method estimates clean conformity scores despite label noise, enabling reliable uncertainty quantification. Next, we study calibration in unsupervised domain adaptation, where a model trained on a labeled source domain is adapted to an unlabeled target domain. Since labeled target data are unavailable, we estimate target-domain accuracy from source performance and domain discrepancies, enabling calibration without target labels. We also consider privacy-preserving settings in which user labels and model outputs must remain protected. We propose a locally differentially private conformal prediction framework that provides valid uncertainty quantification while maintaining privacy guarantees and balancing privacy, computational feasibility, and prediction reliability. Our results bridge calibration theory and practical deployment in safety-critical applications, contributing to reliable, privacy-preserving, and noise-resilient neural network predictions.

Large Vision-Language Models (LVLMs) integrate visual perception with language generation, enabling responses that span image understanding and complex reasoning. However, LVLMs do not just inherit the text-level hallucinations; they also hallucinate against the image, producing fluent responses ungrounded in what they see. This makes LVLM response scoring inherently harder, and our diagnostics show that existing confidence-based metrics adopted from LLMs are insufficient for LVLMs. Specifically, removing the input image barely changes confidence-based selection, suggesting that output-space confidence primarily captures textual plausibility rather than agreement with the image. To address this gap, we propose LookBack, a training-free LVLM response scoring method that augments token likelihood with visual lookback score, a lightweight measure of how strongly each response token refers to image tokens. Across four benchmarks and three models, LookBack consistently improves Best-of-$N$ selection over existing baselines with negligible additional overhead.

Uncertainty quantification (UQ) methods for language models are typically evaluated on single-turn outputs, where uncertainty is attached to one generated answer. For LLM agents, however, the unit of observation is an interactive trajectory, where the model can ask clarifying questions, call tools, update state, and make intermediate decisions whose errors propagate to the final outcome. We study whether three common families of single-turn UQ methods transfer to this setting. Across five LLMs and four multi-turn tool-use datasets from BFCL-v4 and $τ^2$-bench, we evaluate white-box scorers based on action-token probabilities, black-box consistency scorers based on resampled trajectories, and reflexive scorers based on model self-assessment of the trajectory. We find that transfer is often useful but uneven. Token-probability scores are highly sensitive to the choice of aggregator used across turns, reflexive scores provide the strongest low-cost baseline in most evaluated settings, and black-box self-consistency is often the strongest UQ family, with trajectory-equivalence and action-set consistency typically ranking highest among its variants. These results suggest that UQ methods developed for single generations should be revalidated at the trajectory level, with careful attention to the consistency measurement, aggregator choice, and computational budget.

We propose that a model's uncertainty about a token is reflected not only in the breadth of its output distribution but also in whether a confident prediction is \emph{fragile} under perturbation of its attention pathways. We instantiate this as ASMI (Attention-Subnetwork Mutual Information), a training-free estimator that masks attention heads and measures the BALD mutual information among the resulting subnetworks, with a semantic-agreement kernel to discount surface-form disagreement. The signal is not a restatement of output confidence: on grounded QA an out-of-fold test shows it adds error-predictive information beyond single-pass confidence and entropy, concentrated in \emph{confident-but-fragile} predictions, where acting on it roughly halves the retained error of a confidence filter. The distinctness is regime-graded, so ASMI predicts its own domain of applicability, strong where answers are routed through provided context and bounded by design where they are recalled from parametric knowledge. Sem-ASMI reads the signal from a single greedy response, without the stochastic generations the strongest baselines require, and ties or beats Semantic Entropy on ten of the twelve grounded benchmark-backbone settings. Across the same twelve settings, the best ASMI variant, typically the adaptive one reusing the ten samples already drawn for the baselines, ties or leads the strongest baseline in eight, significantly in three under a paired test. On parametric QA all variants revert to or below the zero-cost MSP baseline, exactly as predicted, and the estimates are near-deterministic across reruns. A head-level analysis shows that what tracks this boundary is not the presence of head-level fragility but whether that fragility couples to errors.

Reliable semi-dense matching is essential for modern geometric vision systems. Designed under a coarse-to-fine paradigm, it achieves an optimal balance between performance and computational cost. However, existing methods often struggle to provide well-quantified uncertainties, where catastrophic coarse-assignment failures are ignored, leading to truncated error distributions and severely misjudged geometric estimations. In this paper, we propose a lightweight, post-hoc overall uncertainty estimation framework that introduces a two-component calibrated Laplace mixture model with only 9 learnable parameters. The objective is to explicitly capture both the sharp local refinement noise and the broader tail of coarse-assignment failures. We introduce the Coarse-success posterior Refit (CoRe) method, a geometric refitting module that utilizes the posterior probability of coarse-assignment success as soft correspondence weights. Extensive experiments show that our method consistently improves downstream geometric accuracy across various pretrained-only matchers and robust estimators with minimal computational overhead. Our code is available at https://github.com/khoavpt/Probabilistic-matching.

Deployed vision-language systems often gate their answers on confidence, making confidence robustness relevant to oversight. We study confidence readouts under white-box, image-only attacks constrained to preserve the generated answer byte-identically. Under a reachability assumption, an unmovable readout cannot outperform the answer-string accuracy prior, whose pooled value is 0.617. Independently of that assumption, a uniform amplitude certificate below a measurable threshold guarantees adversarial discrimination above the same floor. Across four vision-language models, three visual question answering benchmarks, five deployed confidence channels and two defense estimators, direct or surrogate-aimed attacks produce itemwise feasible perturbations that refute this uniform certificate in all 84 estimator-by-cell combinations. Coordinated correctness-label-aware attacks drive adversarial discrimination to or below the answer-string floor in all sixty deployed-channel cells, including all fifty-nine that begin above it. Hidden-state interventions and an open-ended text-model activation-space replication show that comparable confidence movement can be induced at the representation level rather than only through adversarial images. None of four tested defense families establishes a robust alternative under the specific evaluation applied to it. In a confidence-gated simulation, a coordinated token-probability attack transferred to a hidden-state gate causes up to 84.8% of previously rejected wrong answers to become accepted. After reweighting to each benchmark's natural correctness prevalence, accepted accuracy falls below the no-gate baseline in eight of twelve cells under transfer and all twelve under a direct gate-aimed attack. Under the studied threat model and budget, confidence is therefore an integrity-sensitive rather than intrinsically robust oversight signal.

Web agents observe a browser through text, pixels, or both, and the choice is usually fixed once for all tasks. We measure six observation modes across eight site-model combinations (cells) on VisualWebArena and WebArena and ask what choosing per task would buy. The modes are complementary: each solves tasks the others miss, they fail in structurally different ways, and the best choice reverses between task sets. The obvious prize, an oracle that picks a winning mode for every task, looks large but is inflated by run-to-run noise: rerunning the same mode on the same tasks changes 12-14% of outcomes, so a second run of a mode already in hand gains about as much as adding a new one. What survives is a cost bound: sending only the tasks no mode solves to the cheapest mode cuts cost by 9.5-30.6% in 8 of 8 cells at unchanged success. We then test five routing policies (picking the mode, deciding when to spend on the strong mode, a zero-cost rule read off the task text, a confidence cascade, and pooled cost tiers), and none robustly beats simply fixing one well-chosen mode; the one exception is a fragile result in our sparsest cell. The central obstruction is that routing supervision is produced at the agent's success rate: the weaker the agent, the fewer labels a router gets, exactly where routing would be most valuable. This limit belongs to today's agents rather than to routing itself. Label supply and routing opportunity rise together (correlation 0.95 across cells), so a stronger agent can overturn the result, and we report the rerun noise bands and the full measurement protocol.

Continuum robots often operate in uncertain environments, where accurate state estimation is essential for safe interactions. Estimate uncertainty is inherently spatially non-uniform: confidence varies depending on where measurements are available. Global estimation accuracy is not always the top priority, but rather achieving sufficient confidence at task-relevant locations along the robot. This extended abstract introduces mechanically reconfigurable sensing enabling uncertainty-shaping in state estimation for continuum robots. We present a concept hardware design demonstrating the feasibility of longitudinal translation of a sensor within a continuum robot. We demonstrate that state estimation confidence can be reconfigured by varying the sensor location, and show a reduction of full-body shape estimation errors when sliding the sensor back and forth over time, compared to a single fixed tip sensor.

Refurbishment-based noisy-label learning mixes an observed label with a model-derived pseudo target, typically using one sample-wise cleanliness score to control both branches. This creates a hidden coupling: reducing trust in the observed label automatically increases trust in the pseudo target. We show that this complementarity can replace one unreliable signal with another because a pseudo target learned from corrupted supervision may reproduce the noise it is meant to correct. Our representation diagnostics provide a consistent account of this mismatch: noisy supervision redirects deeper layers more strongly, whereas shallower relations remain comparatively stable and provide information beyond the loss posterior. We therefore propose TRACE, a Two-Source Reliability Assessment framework for Label Correction and Sample Reweighting. TRACE assesses the observed label using loss fit, shallow relation stability, and prediction agreement, while separately assessing the pseudo target using model confidence. Its source-specific scores control target correction and supervision strength without assuming complementary reliability. Across synthetic and real-world noisy benchmarks, TRACE improves representative refurbishment baselines and yields more reliable pseudo supervision.

Metasignal is an open-source Python package for signal detection theory (SDT) and metacognitive measurement. It implements the 17 metacognitive measures evaluated by Rahnev (2025), together with the reference variables d' (perceptual sensitivity), response criterion c (response bias), and mean confidence. The 17 measures comprise three meta-d' family estimates, meta-d', M-ratio, and M-difference; four nonparametric Type-2 measures, the Type-2 area under the receiver-operating-characteristic curve (AUC2), Gamma, Phi, and delta confidence, together with their eight SDT-normalized ratio and difference forms; and two model-based measures, meta-noise and meta-uncertainty. A single function computes the complete set from trial-level stimulus, response, and confidence arrays. metasignal currently supports binary (two-alternative) discrimination tasks, in which each trial's stimulus and response are coded with exactly two categories. The package also provides a command-line interface, group summaries, bootstrap confidence intervals, permutation tests, optional hierarchical Bayesian models, and information-theoretic measures. metasignal unifies these measures in a single platform to encourage broader metacognition research and adoption in decision-making studies.

Vision-language-action (VLA) policies commonly execute long-horizon mobile manipulation through open-loop action chunks, issuing multiple actions without receiving new high-level visual input. A committed chunk therefore implies how observations should evolve, but accidental deviations can violate this expectation while the remaining actions continue to propagate the error: commit-time policy confidence cannot react to a deviation that occurs after dispatch, and observation-only anomaly scores lack an action-conditioned reference for separating expected effects from unexplained changes. We propose CheckVLA, which verifies execution with a separately trained, frozen action-conditioned world model. A conformally calibrated risk threshold bounds the episode-level probability of an unnecessary first intervention and determines when to intervene, its exceedance controls how strongly the rewritten suffix retains the superseded chunk, latency-aware hard prefixing restricts replacement to actions that remain deployable, and an event-driven keyframe bank preserves evidence of prior progress across repairs. On RoboCasa365, under a common training recipe and a matched invocation budget, CheckVLA attains a 36.1% average success rate against 27.6% for periodic replanning (+8.5 points). At a matched 5% episode-level false-alarm target, action conditioning raises timely recall to 77.9%, against 48.6% for an observation-only control and 37.9% for an action-shuffled control. These simulation results support action-conditioned verification as a way to restore feedback during chunked execution while keeping the repair consistent with inference latency.

Evaluation methodologies for language models increasingly combine multiple signals, from automated metrics and LLM-as-judge ratings to human assessments and benchmark suite results. When these signals are aggregated via averaging, evaluation confidence can then substantially exceed the reliability of the weakest signal: a phenomenon we call trust inflation in evaluation. We argue that evaluation scores should be treated as epistemic claims with three properties: formality (human evaluation provides stronger evidence than an automated metric), scope (a benchmark result applies to the tested distribution, not universally), and validity windows (benchmark results expire as contamination accumulates and distributions shift). Several converging research traditions (chain-of-thought analysis, possibilistic logic, and algebraic theory) establish weakest-link aggregation as the conservative endpoint of a parameterized operator family controlled by a single pessimism parameter. Drawing on those traditions, and on concrete lessons from building an evaluation harness for agentic AI, we propose that evaluation results carry explicit metadata (formality tier, scope declaration, and expiration date) to make their epistemic status transparent. We illustrate the cost of mean aggregation on the public HELM leaderboard: across 54 frontier models on ten scenarios, the top-five models ranked by mean score and by weakest-link are completely disjoint.

Vision-language models (VLMs) are increasingly deployed on consumer hardware where input images are degraded by compression, camera shake, and poor lighting. In such settings, a reliable uncertainty signal matters more than raw accuracy, because it determines when a system should defer rather than answer. We evaluate two small open-weight VLMs -- Qwen2-VL-2B-Instruct and SmolVLM-Instruct -- across six realistic photographic degradations at three severity levels, comparing two confidence signals: the confidence the model states in natural language, and the model's own mean token probability over its generated answer. Across 3,800 predictions, we find a large and consistent gap. Verbalized confidence in Qwen2-VL is almost constant (mean 0.87-0.90 across all conditions) and detects its own errors at chance level (AUROC 0.39-0.75, typically ~0.50), while internal token probability from the same model separates correct from incorrect answers with AUROC 0.92-0.99. In SmolVLM, verbalized confidence proved largely unobtainable: across three prompt templates, only one of five pilot attempts produced a parseable confidence value, while internal probability again yielded above-chance error detection (AUROC 0.54-0.92). Both models fail in the same place: under severe underexposure, accuracy collapses (0.99->0.22 for Qwen2-VL, 0.97->0.42 for SmolVLM) while both confidence signals barely move, and internal error-detection falls to chance. We conclude that small VLMs encode usable self-knowledge that their verbalized output does not express, that internal probability is therefore the better deferral signal in constrained deployment, and that neither signal should be trusted under severe low-light conditions.

We study the problem of sequentially evaluating a new large language model (LLM) on a fixed question set using historical performance data from prior LLMs. Our goal is to construct a confidence sequence (CS) for the model's capability on this question set and to design active querying rules that shrink the CS width as quickly as possible. For CS construction, we invert a family of test supermartingales and focus on two representative approaches: a reverse information projection (RIPr)-based approach and a testing-by-betting-based approach. We first study these approaches under an oracle setting, and demonstrate the oracle optimality of the RIPr-based construction. We then propose a growth-oriented querying rule that aims to maximize the worst-case one-step expected log-increment over the endpoints of the current CS. In practice, we build these test supermartingales and the querying rule on predictions of question-level correctness learned from historical data. We then analyze the shrinkage behavior of the resulting CSs and identify two key factors that slow the shrinkage rate of CSs: accumulated prediction mismatch and the spikiness of the querying distribution. Finally, motivated by this analysis, we propose several mixture querying rules that combine growth-oriented querying, prediction refinement, and uniform exploration, trying to mitigate the effects that slow the shrinkage rate. We provide experiments comparing different querying rules for the RIPr-based and testing-by-betting-based CSs across several synthetic testing datasets. Interestingly, we observe that the simplest querying rule, uniform sampling, can sometimes outperform more adaptive querying rules for both methods.

LLM agents act in external environments where each action changes the state that later decisions condition on, and where a single wrong step can waste interaction budget or trigger irreversible side effects long before the final failure is observed. Reliable deployment therefore requires \emph{step-level confidence estimation}: a calibrated probability that each proposed action is productive, available \emph{before} the action is executed. Existing LLM confidence estimators are designed to score a response from the given prompt, but agent confidence also depends on execution consequences: whether similar actions in similar situations actually advanced the task after the environment responded. We introduce the \method (\methodshort), a self-evolving critic framework in which an LLM critic accumulates evidence from its own past judgments and their observed consequences. After each trajectory, a hindsight LLM that sees the full execution feedback votes on whether each step was productive. The resulting pseudo-labels populate a memory bank from which related productive and unproductive experiences are retrieved into the critic's prompt whenever a similar step recurs. \methodshort requires no training and uses no ground truth step labels. Across three agent benchmarks and three critic backbones, \methodshort attains the best calibration (ECE and Brier) and ranking (AUC) in every dataset--critic combination, reducing ECE by up to $54\%$ relative to the strongest training-free baseline.

We present our submission to the QANTA 2026 shared challenge at the ICML 2026 Workshop on Efficient Multimodal Question Answering (EMM-QA). Quanta evaluates multimodal quizbowl systems that answer pyramid-style questions from incrementally revealed text and accompanying images while operating under realistic efficiency constraints. The challenge consists of two distinct tasks: Tossup questions, which require deciding when to answer under uncertainty, and Bonus questions, which emphasize accurate answer selection and human adoption. To address these differing objectives, we develop a task-specific two-agent architecture. Our Tossup agent utilizes a GPT-4o-mini-class model (referred to as GPT-4.1-mini in the competition logs) with confidence-calibrated answering and a domain-specific numeric reasoning policy that reduces overconfident predictions from isolated quantitative clues. Our Bonus agent uses GPT-4o-class model (referred to as GPT-4.1) with leadin-aware reasoning, structured relational reasoning, and multimodal evidence integration to improve exact answer selection. Rather than relying on a retrieval pipeline or model ensembles, our approach emphasizes efficient reasoning policies and confidence calibration within a hosted-only environment. Our system achieved the highest overall leaderboard score of 0.402, including a Tossup score of 0.238 and a Bonus Effect score of 0.164. The results demonstrate that lightweight, task-specific reasoning strategies can provide strong performance on resource-constrained multimodal question answering benchmarks.

Deep models for retinal optical coherence tomography (OCT) classification report high accuracy but rarely report whether their confidence can be trusted -- a gap that matters when a wrong-but-confident reading delays sight-saving treatment. We pair a hybrid convolutional-Transformer encoder with a gradient-boosting (XGBoost) classification head and a three-part clinical safety layer: confidence calibration, out-of-distribution (OOD) rejection, and per-prediction uncertainty flagging. On four-class OCT (84,495 scans) the model reaches 95.4% accuracy while cutting calibration error twelve-fold (expected calibration error, ECE = 0.0024), so the confidence it reports tracks its true accuracy. To our knowledge this is the first OCT classifier to validate all three safety mechanisms jointly, with public weights and reproducible multi-seed evaluation.

Reliable confidence estimation is essential for deploying large language models (LLMs) in confidence-aware systems, where downstream decisions such as retrieval, tool use, and adaptive computation depend on accurately estimating answer reliability. Existing approaches, however, largely treat confidence as a property of completed responses, overlooking how confidence-related information evolves throughout the answering process. In this work, we investigate confidence from a temporal perspective by comparing pre-solution Feeling-of-Knowing (FOK) and post-solution Judgement-of-Learning (JOL) confidence estimates across frontier and open-source LLMs. We show that post-solution confidence is consistently better calibrated and more discriminative than pre-solution confidence, while linear probes trained on hidden representations recover substantially richer confidence-related information than models explicitly verbalise. Building on this observation, we introduce future confidence distillation, which trains predictors operating on pre-solution hidden representations using teacher confidence estimates produced by post-solution correctness probes. Despite requiring only pre-solution representations for inference, distilled predictors recover much of the calibration improvement achieved by post-solution confidence, remain highly sample efficient, and transfer across datasets within the same domain. Together, our findings demonstrate that confidence-related information evolves throughout the answering process and can be anticipated before answer generation is complete, enabling significantly more reliable yet low-cost confidence estimation.

In this paper, we consider the setting where large language models (LLMs) are trained using reinforcement learning (RL) to simultaneously improve reasoning accuracy and verbalize its confidence. Our reward scheme uses two functions for rewarding confidence verbalized by the LLM: one when the LLM is correct and a different one when the LLM is incorrect. With a poorly designed reward scheme, the LLM may be incentivized to answer incorrectly so that it can be confident that its answer is indeed incorrect, a phenomenon that we call confidence reward hacking. We propose the concept of non-hackable confidence reward schemes and define a spectrum of such reward schemes for RL confidence calibration training in LLMs. We demonstrate that selective confidence reward hacking can occur in practical datasets with reward schemes that are not designed to be non-hackable. We also demonstrate that the reward scheme with the best calibration to accuracy tradeoff depends on the dataset and the application, and propose using the reward scheme as a hyperparameter to optimize the tradeoffs in accordance to what is important for the application. The code of our experiments is available in https://anonymous.4open.science/r/rl-confidence-calibration-9ED4/README.md.

Calibration evaluates whether a model confidence aligns with its empirical accuracy. Existing studies often compare the calibration of different large language models using global calibration metrics such as Expected Calibration Error and Brier Score. We begin by showing, both theoretically and empirically, that such comparisons are confounded by differences in model accuracy. For fairer cross-model comparison, we then propose ACE, an accuracy-controlled evaluation framework with three complementary views: Instance-Aligned, Distribution-Aligned, and Candidate-Aligned calibration. Across multiple benchmarks, model families, and confidence elicitation methods, we use ACE to study two practically important comparison axes, small versus large models and thinking versus non-thinking models. We find that many previously reported calibration advantages under raw global metrics weaken substantially after accuracy control. We also find that ranking reversal is frequent: models favored by raw metrics often cease to be favored once accuracy is controlled. Our results show that raw global calibration metrics are not robust for cross-model comparison, and that fair calibration comparison requires accuracy-aware evaluation.

LLM agents carry conclusions across steps and sessions in compressed memory, and memory products (e.g., mem0, LangMem) rewrite conversation into stored "facts" that later steps trust. We show this rewriting manufactures confidence: across our constructed agent settings, a casual, hedged remark becomes a confident, dated assertion the agent then obeys like a verified fact, granting every above-clearance request it faces. No attacker is needed: a role that was true once and never corrected is stored as a flat fact and acted on like a deliberate injection. We then isolate what the agent responds to. It is not the source: attributed, unattributed, and even forged "system of record" claims all grant alike. It is the confidence of the phrasing. A hedge is discounted, a flat assertion is obeyed, and this holds with no special keyword. Not all hedges are equal, though: the evidential register is the least-discounted, with "reportedly" obeyed like a flat assertion on most models. The obvious fixes fail. A passive "unverified" tag is ignored, and an active "do not trust this" instruction escalates even correct memory, so it is safe only by refusing to decide. The real fix lives in the store: keep the tentative phrasing rather than upgrade it. But that is hygiene, not a defense against an attacker who can simply write a confident lie. The deployable lesson is narrower and constructive: a single load-bearing memory is the hazard, and one redundant source restores correct decisions. We release the harness and demonstrations.

Online inspection images of final optics in high-power laser facilities contain pseudo-damage sites that closely resemble true damage sites. Determining the authenticity of online-detected sites is therefore difficult and requires accurate matching to offline ground-truth sites. However, this matching remains highly challenging due to limited match-discriminative features, local geometric distortions, and numerous distractor sites. Existing matching models mainly suppress distractors implicitly through loss-function supervision. We propose a confidence-feedback-weighted graph matching network that requires only damage-site centroid coordinates as input. It estimates node matchability confidence from each round of matching scores and feeds it back as a reliability weight to guide subsequent edge-feature aggregation, thereby suppressing distractor propagation and enhancing cross-graph discriminability. Within this framework, a geometric consistency constraint calibrates spurious high-confidence matchability estimates, while a hard-example mining loss improves discrimination between structurally similar sites. Experiments on our Complex-Scene dataset show that the proposed method achieves a matching F1-score of 96.36$\%$ with robust and efficient performance.

Can a statistically significant, large-effect-size finding in computational social science be entirely an artifact of the measurement instrument? We present a case where the answer appears to be yes. Analyzing 85 interviews across four public intellectuals (2016--2026), we find a robust negative-affect/emphatic-certainty lexical co-occurrence pattern under keyword-based scoring ($r = 0.72$--$0.93$, $p < 0.01$ for all four speakers). Replacing keyword counting with LLM-based zero-shot semantic classification on the complete diarized corpus (32,625 sentences) dramatically reduces this correlation: Dalio's $r = 0.851$ drops to $r = 0.206$, with two speakers showing negative $r(\text{neg}, \text{emphatic})$ and one showing null. In contrast, the LLM reveals a strong negative-hedging coupling across speakers -- Rogoff's $r(\text{neg}, \text{hedged}) = 0.875$ ($p = 0.001$) and Zeihan's $r(\text{neg}, \text{hedged}) = 0.722$ ($p = 0.008$) -- consistent with the conventional expectation that pessimistic discourse attracts hedging, not certainty. Sentence-level error analysis traces this discrepancy to three structural failure modes in keyword lexicons -- syntactic blindness, polysemy blindness, and categorical absence -- illustrated through cases where keyword counting inverts semantic meaning (e.g., ''never absolutely totally confident'' scored as high-certainty). We argue that keyword lexicons measure a universal lexical co-occurrence tendency -- negative discourse naturally attracts emphatic vocabulary -- that is orthogonal to, and can systematically invert, rhetorical stance. Treating keyword counts as measurements of epistemic certainty is a category error: a finding that appears to be about a speaker's psychology may be entirely about the counting of words.

Markov decision processes (MDPs) are a classic model of decision making under uncertainty, exhibiting both non-deterministic choice as well as probabilistic uncertainty. Traditionally, exact knowledge of the underlying probabilities is assumed. However, this often is unrealistic, e.g.\ when modelling cyber-physical systems or biological processes. Here, statistical methods provide a way towards obtaining meaningful guarantees. The classical approach is to gather samples in the MDP, use these to draw statistical conclusions about the transition probabilities, and from there obtain bounds on the true value; then, if these bounds are too broad, repeat. However, existing implementations of this approach are either subtly incorrect or sub-optimal, and quite often both. We present several \emph{confidence sequences}, which are specifically designed for such \enquote{online} settings, implement all of them in an efficient tool, and show their practical applicability. In particular, we show that they outperform classical \enquote{union-bound} style approaches, and overall our implementation requires 50x less samples on average than previous state of the art.

Pruning Large Language Models (LLMs) reduces memory and inference costs by removing parts of the network, producing smaller models that retain most of their accuracy. As attention layers are the most resource-intensive parts of LLMs, pruning them is a promising compression strategy. Prior work shows that up to 33% of attention layers can be pruned with minimal accuracy loss. Nevertheless, the impact of attention pruning on model interpretability, specifically faithfulness and confidence calibration, remains unstudied. To address this gap, we study how pruning attention layers affects explanation faithfulness and confidence calibration across five LLMs and eight datasets. While the pruned models often maintain high accuracy, we find that their faithfulness and calibration often degrade. Notably, faithfulness and calibration can fluctuate significantly, even when accuracy remains stable, highlighting a misalignment between model confidence, interpretability, and accuracy. Our findings suggest that layer pruning can affect LLMs' interpretability and reliability in ways not captured by accuracy and efficiency measures alone. We recommend including explainability and calibration metrics when evaluating pruned models.

Confidence calibration in large language models (LLMs) is commonly evaluated by comparing predicted confidence with observed accuracy. However, such approaches do not model item difficulty, making it difficult to interpret discrepancies and to determine whether model confidence reflects genuine self-assessment or is merely a byproduct of the response generation process. To address this, we adopt a Rasch model-based latent ability framework and a metacognitive perspective, and propose Latent Confidence Alignment Error (LCAE) to measure the consistency between model self-assessment and the latent error probability implied by model ability and item difficulty. We further incorporate item difficulty as an external signal with a reasoning mechanism. Experiments on a medical-domain dataset with 20 models show that the proposed approach improves self-assessment quality without affecting model ability, and reveals an association between reliability and inference cost.

Deep neural networks often produce poorly calibrated confidence estimates, overstating their certainty even when predictions are incorrect. Temperature Scaling remains the most widely used posthoc calibration method due to its simplicity and effectiveness, yet its global, uniform rescaling of logits fails to correct the highly heterogeneous structure of miscalibration observed across the confidence spectrum. In particular, the largest correctness confidence discrepancies arise in different quantile regions depending on the setting, low confidence predictions, where uncertainty matters most, tend to exhibit the largest correctness confidence discrepancies, which standard TS leaves largely unaddressed. We introduce Quantile Adaptive Temperature Scaling (QaTS), a simple and efficient post hoc calibration method that adapts the temperature as a function of a predictions empirical confidence quantile. By mapping confidences into the quantile space, QaTS normalizes the calibration problem, makes the structure of miscalibration explicit and enables a monotone temperature function that adapts across quantiles while leaving well calibrated high confidence predictions largely unchanged. preserving high confidence behavior. This quantile aware formulation aligns naturally with a reparameterized Expected Calibration Error (ECE) objective and yields a sample wise temperature that is robust across a variety of challenging scenarios, such as class imbalance and distributional shifts. Across a broad range of datasets, architectures, evaluation scenarios and diverse tasks, QaTS consistently, and substantially, outperforms state of the art post hoc calibration methods, delivering more reliable and trustworthy confidence estimates without modifying model predictions.

Multimodal Large Language Models (MLLMs) show great potential in medical tasks, but their elicited confidence often misaligns with actual accuracy, potentially leading to misdiagnosis or overlooking correct advice. This study presents the first comprehensive analysis of the relationship between accuracy and confidence in medical MLLMs. It proposes a novel method that combines Multi-Strategy Fusion-Based Interrogation (MS-FBI) with auxiliary expert LLM assessment, aiming to improve confidence calibration in Medical Visual Question Answering (VQA). Experiments demonstrate that our method reduces the Expected Calibration Error (ECE) by an average of 40% across three Medical VQA datasets, significantly enhancing MLLMs' reliability. The findings highlight the importance of domain-specific calibration for MLLMs in healthcare, offering a more trustworthy solution for AI-assisted diagnosis.

Multimodal emotion-cause pair extraction (MECPE) requires reliable pair confidence over candidate pairs. Existing pair scorers commonly use pair-level cross entropy over valid candidates, which treats links mostly independently. This leaves the relative confidence geometry among competing causes under-constrained, allowing gold pairs to stay close to hard negatives or rely on incidental non-gold context. We study this vulnerability as pair-confidence brittleness and propose RPCL (Robust Pair Confidence Learning), a training-only framework for pair-confidence learning. RPCL encourages pair confidence to be both discriminative and stable: gold pairs are separated from row-wise hard negatives through a confidence-difference margin constraint, and clean pair predictions are aligned with predictions from a corrupted view where non-gold contextual utterance representations are partially corrupted. The original clean pair scorer and decoding pipeline are used unchanged at inference time. On ECF, MECAD, and MEC4, RPCL improves the three-seed mean Pair F1 over a matched base model by 2.58 to 2.83 percentage points in the full text-audio-video setting, and improves mean Pair AUPRC on all three datasets. Diagnostic analysis further shows larger gold-negative confidence gaps and lower margin-violation severity. These results suggest that explicitly shaping pair confidence is an effective training strategy for MECPE.

Machine learning is increasingly applied to accelerate the discovery of novel materials by exploring large compositional and structural design spaces. Yet, the scarcity of high-quality data and the frequent need for out-of-distribution prediction introduce substantial uncertainty, making the assessment of model reliability essential. In this work, we investigate uncertainty quantification as a means to evaluate model confidence in the context of permanent magnet research. In a first study, we benchmark classical and modern machine learning models for predicting intrinsic magnetic properties, focusing on the quality of their uncertainty estimates. We apply Gaussian negative log-likelihood loss and dropout-based Bayesian approximation as practical strategies for estimating predictive uncertainty. In a second study, we transfer these architectural features for uncertainty estimation to a more complex task: predicting coercivity from microstructural information using a graph neural network. Together, these studies demonstrate that uncertainty quantification not only enhances the trustworthiness of predictions but is also transferable across different modeling tasks.

Evaluating reasoning quality in multi-agent LLM systems is challenging, especially for open-ended tasks without reference answers. We investigate whether intrinsic confidence signals, token-level log-probabilities from decoding, can predict reasoning quality as assessed by LLM-as-judge evaluation. Using a debate-based essay scoring framework, we compare confidence proxies against rubric-based judge scores across two ASAP essay sets. We find that early-token confidence, particularly within the first few generated tokens, is consistently the strongest predictor of reasoning quality, outperforming full-sequence statistics. Analysis of log-probability trajectories shows that the opening phase of generation is the most heterogeneous and therefore most informative. We also observe a systematic asymmetry between agent roles, with stronger alignment between confidence and quality for supportive reasoning than for adversarial critique. These results suggest that early decoding dynamics provide a lightweight and effective signal for estimating reasoning reliability in multi-agent LLM systems.

Qualitative product feedback can reveal nuanced user experiences, but its implicit sentiment is difficult to measure. This paper presents a scalable and interpretable framework that uses large language models (LLMs) to quantify product desirability from such data. Using two Product Desirability Toolkit (PDT) datasets from ZORQ and CARMA comprising 106 respondent term groupings with gold-standard human annotation, zero-shot continuous numerical sentiment scoring and categorical sentiment classification are evaluated without relying on explicit review scores. Across the datasets, LLMs generated numerical sentiment scores directly from qualitative responses and closely matched expert labels, achieving Pearson correlations up to 0.97 and classification accuracy up to 94%. LLMs maintained robustness even when handling data presented in multiple forms and consistently expressed high confidence. In contrast, lexicon-based and transformer baselines did not produce statistically significant results. Among the models tested, GPT-4o-mini achieved performance comparable to larger models at 94% lower cost, supporting scalable deployment. The framework also incorporates model confidence ratings and human-readable rationale explanations (xAI), improving interpretability, transparency, and trust while supporting practical use in product satisfaction assessment. In general, using the PDT tool as a survey method along with a cost efficient LLM for sentiment analysis has the potential to provide for product evaluation with results that are rich in terms of sentiment scores (both numerical and classified sentiment) and in terms of the high-level user impressions of the product that can be used to identify ideas for product development and improvement, as well as marketing ideas for target audiences.

Large language models often express high confidence in answers that are wrong. Standard calibration remedies typically act globally or at the score level, reducing unwarranted confidence but also risking erosion of warranted confidence on correct answers. We introduce Probe-Conditioned Head Intervention (PCHI), an inference-time method that uses a frozen probe to detect likely wrong-but-confident responses and conditionally rescales downstream attention-head outputs during confidence generation. On Qwen3-4B-Instruct solving OpenMathInstruct problems with a structured binary confidence field, readout-token PCHI converts 82.2% of originally wrong-yes confidence readouts to $\texttt{no}$, while a joint intervention across upstream confidence-template tokens reduces ECE from 21.9% to 9.2% and damages only 5.1% of originally correct-yes readouts. The readout-token effect also appears on Gemma3-4B, though upstream interventions are weaker and more mask-dependent. These results show that verbalized overconfidence can be selectively reduced through conditionally applied internal intervention, partially decoupling the suppression of unwarranted confidence from the loss of warranted confidence.

Reliable uncertainty communication is critical to the trustworthiness of LLMs, yet faithful calibration (FC)--the alignment between models' intrinsic and (linguistically) expressed confidence--is a persistent failure mode. This challenge is key for large reasoning models (LRMs), whose extended reasoning traces are often interpreted by users as evidence of deliberation, competence, and confidence. Despite the importance of FC and wide usage of LRMs, the extent to which LRMs can faithfully express their confidence remains poorly understood. Moreover, the prevailing paradigm to measure FC does not generalize well to the long chain-of-thought outputs generated by LRMs, which tend to lack clear step boundaries, involve inconsistent step structure, and encode complex conditional dependencies throughout the trace--complicating estimation of intrinsic confidence. To address this challenge, we introduce a novel framework to systematically quantify FC of LRMs. Our framework analyzes linguistic decisiveness relative to three sources of internal uncertainty, based on token probabilities, hidden states, and sampled response consistency. We also devise a prefix-conditioned sampling approach to control for conditional and structural variation across traces. Applying our framework to a diverse suite of leading models, datasets, and prompts, we find that faithful confidence expression is a significant challenge for LRMs. Reasoning behaviors do not automatically translate to improved FC, and prompt interventions for non-reasoning models do not improve faithfulness in the reasoning setting. Different confidence estimators further produce divergent assessments of the same traces, revealing fragility in prior evaluation methodologies. Taken together, our work establishes FC as a distinct reliability and alignment target for LRMs, particularly as such systems are increasingly deployed in high-stakes contexts.

Conversational tutoring agents have been shown to improve learning engagement and student outcomes, and large language models (LLMs) are increasingly used in these systems to provide scalable, personalized feedback. However, LLMs may perpetuate or amplify stereotypical social biases, posing particular risks in educational settings. In this study, we evaluate LLMs in conversational tutoring scenarios to identify high-confidence social biases, instances where models are unable to identify biased judgments in tutoring conversations while maintaining strong confidence in their assessments, potentially affecting their reasoning and the feedback they provide to learners. We present a new dataset generation method that enables bias evaluation under naturalistic instructional conditions by regenerating student-AI tutor interactions and introducing turns with controlled bias derived from a benchmark dataset. Using this data, we assess multiple LLMs' ability to detect stereotypical biases and analyze the confidence and reasoning underlying their responses through computational and human evaluations. We find that bias detection is substantially more challenging in conversational tutoring contexts than in benchmark-based evaluations, and that state-of-the-art LLMs are overconfident in their incorrect assessments of stereotypical bias statements. Moreover, model confidence strongly influences reasoning and feedback, highlighting the risks of overconfident, biased behavior in LLM-based tutoring agents. We conclude by discussing implications, mitigation considerations, and directions for future research.

LLMs' linguistically expressed confidence should faithfully reflect their intrinsic uncertainty. While recent work shows LLMs struggle to use epistemic markers (e.g., "it is likely...") in a human-aligned fashion, it remains unclear whether models can apply their own linguistic confidence framework to associate markers with specific confidence levels in a stable and generalizable way, and how contextual features impact this ability. We conduct the first systematic study of this question, formalizing marker internal confidence (MIC) as the estimated intrinsic confidence a model associates with a specific epistemic marker in a given task domain. We present 7 metrics to evaluate the stability of MICs within and across distributions. Applying our analysis framework to diverse models and tasks, we find that LLMs remain faithfully miscalibrated even under model-centric interpretation of marker meanings, struggling to differentiate markers by internal confidence across distributions despite preserving a somewhat consistent ranking order across tasks. This supplies critical, complementary evidence to existing work toward a holistic understanding of faithful calibration in LLMs, emphasizing the need for more aligned and stable marker use to improve trustworthiness and reliability.

AI systems are fallible, and humans can make mistakes in deciding whether to trust AI over their own judgment. Thus, improving human-AI collaboration requires understanding when, why, and how humans decide to rely on AI. We study two distinct reliance decisions: the delegation choice -- deciding when to let AI act autonomously without knowing its output, and the adoption choice -- evaluating AI suggestions and deciding how to use them. Both of these decoupled reliance patterns shape collaboration, but prior work rarely studies them together in realistic settings with the same users. We address this gap by studying collaborative human--AI teams competing in a question-answering game in which humans can choose when and how to work with AI agents to win. Our 24 matches pair 23 expert humans with 16 AI agents, capturing 387 delegation and 1440 adoption decisions. While human--AI collaboration performs better than either AI or humans alone, humans make suboptimal collaboration decisions, both under-relying on correct AI suggestions (3.9% of opportunities missed) and over-relying when AI misleads them (1.7%). Both parties contribute wrong answers: reported model confidence is near chance when humans and AI disagree, while confirmation bias drives higher under-reliance (64.5%) when an AI suggestion agrees with humans' initial incorrect answer. To close this gap, we recommend calibrated confidence, evidence-grounded explanations, and mechanisms that help users refine trust.

Large language models (LLMs) are prone to hallucinations, i.e., statements unsupported by the input or training data, hindering reliable deployment. In parallel, numerous uncertainty estimation (UE) methods have been proposed to quantify model confidence and are often implicitly treated as proxies for model failure. However, the relationship between uncertainty and hallucinations remains insufficiently characterized. We present a systematic empirical study of the association between uncertainty estimators and hallucinations in LLMs. Rather than assuming this association, we evaluate directly when and to what extent it holds. We consider a diverse set of uncertainty estimators, including information-theoretic, sampling-based, and reflexive estimators, and examine their behavior across hallucination settings. Our experiments cover both intrinsic hallucinations (violations of input faithfulness) and extrinsic hallucinations (unsupported claims relative to training data), using four complementary benchmarks, including RAGTruth and HalluLens. We find that the association is highly variable and often weak, depending on the hallucination type and the LLM under evaluation. These results challenge the use of uncertainty as a direct signal of hallucination and clarify when it provides actionable information.

Semi-supervised learning has become a dominant paradigm for reducing annotation costs. However, we argue that the current progress is clouded by a twofold overconfidence problem. Algorithmically, mainstream pseudo-labeling frameworks often conflate prediction confidence with uncertainty, leading to severe confirmation bias. Strategically, since multiple benchmark datasets lack dedicated validation sets, some studies use the test set for validation as well, leading to inflated performance estimates. Subsequent methods, compelled to employ the same strategy to surpass reported SOTA, trigger an arms race of overfitting. This raises concerns that the impressive numerical gains in the community may reflect overfitting rather than genuine progress. Thus, we propose a tri-space calibrated segmentation framework founded on a principled dual-axis reliability assessment engine. It explicitly decouples confidence from uncertainty and uses this signal to detect and correct confirmation bias across feature, probability, and image spaces in a collaborative manner. Across three benchmark datasets, TCSeg consistently delivers strong performance under existing evaluation protocols. More importantly, we advocate that the community report final-checkpoint results under multiple-run protocols, thereby establishing more rigorous benchmarks with a more realistic perspective. Code will be available: github.com/DirkLiii/TCSeg.

Confidence-weighted routing, selective abstention, and ensemble weighting all assume that a model's stated confidence is informative about its capability on the question being asked. They presume functional metacognition, the capacity to assess one's own capabilities, without exercising them. Aggregate calibration is well studied, with mixed results, but the underlying structure of elicited confidence is less well understood. We decompose binary confidence judgements from 20 frontier Large Language Models (LLMs) across six benchmarks using tetrachoric factor analysis paired with pairwise calibration, asking whether two models that differ in confidence also differ in performance. On factual recall and information retrieval benchmarks the cross-model confidence matrix is approximately rank-one and a single dominant factor captures most of the latent variance. Models retrieving facts share an item-level difficulty axis and differ mainly in their decision thresholds along it. Across all benchmarks the relationship between confidence and performance collapses once items that all models agree on are removed. Inter-model pairwise calibration is small even where statistically significant, and what remains shrinks to nothing once base-rate differences along the shared factor are controlled for. Mathematical reasoning is the apparent exception, but this turns out to be a confound where reasoning models answer questions about their confidence by trying to solve them in their chain of thought, bypassing the sub-symbolic self-knowledge we seek to measure. We find no evidence for significant verbalised individuated metacognition in any tested domain.

We present a method for dynamic quantitative assurance that enhances static safety cases with continuous, runtime-driven confidence updates. The method quantifies and propagates confidence across the development lifecycle by integrating design-time evidence and windowed runtime Safety Performance Indicators (SPIs) within a single Subjective Logic (SL)-based assurance case. At runtime, SPI evidence is continuously evaluated, and targeted claims are updated using a rule that increases confidence in the absence of violations and imposes prompt penalties when violations occur. This design prioritizes safety-relevant responsiveness over exact classical Bayesian posterior updates. We demonstrate the method using a simulation-based construction zone assist function, focusing on an ML-based construction cone detection component, and show how confidence evolves as SPI evidence is observed in operation.

Large language models (LLMs) have become increasingly capable of following instructions and complex reasoning, making prompting a flexible interface for adapting models without parameter updates. Yet prompt design remains labor-intensive and highly sensitive to formatting, phrasing, and instruction order, motivating automated prompt optimization methods that reduce manual effort while preserving inference-time flexibility. However, existing methods often search over prompt candidates or use fixed critique-refine pipelines driven by individual examples or small batches, limiting their ability to capture systematic error patterns and make targeted edits grounded in failure history. We propose Reflective Prompt Tuning (RPT), a framework that uses LLM function calling to simulate the iterative workflow of human prompt engineers. An LLM optimizer calls a diagnostic function that evaluates the target model over an entire optimization set, summarizes recurring failure modes, and returns a structured diagnostic report. The optimizer uses this report, together with an accumulated memory of prior reports, to revise the prompt for the next iteration. RPT further supports confidence-aware optimization by using calibration signals in diagnostic feedback and final prompt selection. Across three reasoning tasks, RPT improves over initial prompts by up to 12.9 points, remains competitive with state of the art, and improves confidence calibration. Our analyses show that RPT is especially effective on multi-hop and mathematical reasoning, producing targeted prompt revisions that align with diagnosed failure patterns and lead to gains in task performance and calibration.

Confidence calibration for classification models is vital in safety-critical decision-making scenarios and has received extensive attention. General confidence calibration methods assume training and test data are independent and identically distributed, limiting their effectiveness under covariate shifts. Previous calibration methods under covariate shift struggle with class-wise or canonical calibrations and often rely on unstable importance weighting when density ratios are large or unbounded. Given the above limitations, this paper rethinks confidence calibration under covariate shifts. First, we derive a necessary and sufficient condition for confidence calibration under covariate shifts, named Expectation consistency condition, which reveals covariate shifts do not necessarily lead to uncalibrated confidence and provides a weaker condition for confidence calibration than global covariate distribution alignment. Then, utilizing Expectation consistency condition, this paper proposes an unsupervised domain adaptation loss to calibrate confidence of the target domain, named Expectation consistency loss (ECL), which is compatible with canonical calibration, class-wise calibration, and top-label calibration. Third, we prove that computing ECL loss has the same sample complexity as Expected Calibration Error (ECE) and provide a theoretically grounded mini-batch trainable scheme for ECL loss. Finally, we validate the effectiveness of our method on both simulated and real-world covariate shift datasets.

Uncertainty Quantification (UQ) is widely regarded as the primary safeguard for deploying Large Language Models (LLMs) in high-stakes domains. However, we argue that the field suffers from a category error: mainstream UQ methods for LLMs are just unsupervised clustering algorithms. We demonstrate that most current approaches inherently quantify the internal consistency of the model's generations rather than their external correctness. Consequently, current methods are fundamentally blind to factual reality and fail to detect ``confident hallucinations,'' where models exhibit high confidence in stable but incorrect answers. Therefore, the current UQ methods may create a deceptive sense of safety when deploying the models with uncertainty. In detail, we identify three critical pathologies resulting from this dependence on internal state: a hyperparameter sensitivity crisis that renders deployment unsafe, an internal evaluation cycle that conflates stability with truth, and a fundamental lack of ground truth that forces reliance on unstable proxy metrics to evaluate uncertainty. To resolve this impasse, we advocate for a paradigm shift to UQ and outline a roadmap for the research community to adopt better evaluation metrics and settings, implement mechanism changes for native uncertainty, and anchor verification in objective truth, ensuring that model confidence serves as a reliable proxy for reality.

Generative models can produce nonsensical text, unrealistic images, and unstable materials faster than simulation or human review can absorb; without per-sample confidence, trust erodes. Existing fixes run $k$ ensembles or stochastic trajectories at $k\times$ compute, measuring variability between models, not model confidence. We propose Flow Matching with Confidence (FMwC). FMwC injects input-dependent multiplicative noise at selected layers, propagates its variance through the network in closed form, and integrates it along the ODE trajectory, yielding a per-sample confidence score at standard sampling cost. The score supports multiple uses: filtering improves image quality and thermodynamic stability of crystals; editing rewinds trajectories to the points where the model commits and redirects them; and adaptive stepping concentrates ODE compute where the flow is ambiguous. We find that the confidence score correlates with the magnitude of the divergence of the learned velocity field, which gives us a window to understand the generative process, opening up surgical forms of guidance that target the moments that matter, new sampling algorithms and interpretability of generative models.

Calibration is commonly evaluated by comparing model confidence with its empirical correctness, implicitly treating reliability as a function of the confidence score alone. However, this view can hide substantial structure: models may be systematically overconfident on some kinds of inputs and underconfident on others, causing global reliability diagnostics to obscure localised calibration failures. To address this, we formulate the problem of discovering hidden miscalibration regimes without assuming access to predefined data slices. We define the corresponding miscalibration field and propose a diagnostic framework for estimating it. Our approach learns a calibration-aware representation of the input space and estimates signed local miscalibration by kernel smoothing in the learned geometry. Across four real-world LLM benchmarks and twelve LLMs, we find that input-dependent calibration heterogeneity is prevalent. We further show that the discovered fields are actionable: they support local confidence correction and reduce calibration error in systematically miscalibrated regions where confidence-based methods such as isotonic regression and temperature scaling are less effective.

The Metacognitive Probe is an exploratory five-task, 15-slot diagnostic that decomposes an LLM's confidence behaviour into five behaviourally-distinct dimensions: confidence calibration (T1-CC), epistemic vigilance (T2-EV), knowledge boundary (T3-KB), calibration range (T4-CR), and reasoning-chain validation (T5-RCV). It is evaluated on N=8 frontier models and N=69 humans. The instrument is motivated by Flavell (1979) and Nelson and Narens (1990) but operates on observable confidence-correctness alignment; it is not a validated cross-species metacognition scale, and the pre-specified human developmental hypothesis was falsified. Composite benchmarks (MMLU, BIG-Bench, HELM, GPQA) ask whether a model produces a correct response. They are silent on whether the model knows when its response is wrong. A model can score 80 on a composite calibration benchmark and still be wildly overconfident in narrow pockets the aggregate cannot surface. The Metacognitive Probe surfaces those pockets. Our headline is a 47-point within-model dissociation in Gemini 2.5 Flash: panel-best within-task calibration (T1-CC = 88; Spearman rho = +0.551, 95% CI [+0.14, +0.80], p = 0.005) and panel-worst cross-task difficulty prediction (T4-CR = 41; sigma_conf = 1.4 across twelve factoids).

Confidence sequences based on test martingales provide time-uniform uncertainty quantification for the mean of bounded IID observations without parametric distributional assumptions. Their practical efficiency, however, depends strongly on the choice of martingale updates, and many existing constructions do not exploit prior information about plausible data-generating distributions or mean values. We propose a Bayes-assisted framework that uses a Bayesian working predictive model to adaptively construct confidence sequences. For each candidate mean and time point, the predictive distribution selects, among valid one-step martingale factors, the update maximising predictive expected log-growth; validity is therefore preserved even when the prior or working model is misspecified. We prove that if the predictive distribution is Wasserstein-consistent, the resulting procedure is asymptotically log-optimal, matching the per-sample log-growth of an oracle procedure with access to the true distribution. We instantiate the framework using robust predictives based on Dirichlet-process mixtures and Bayesian exponentially tilted empirical likelihood. Experiments on synthetic data, sequential best-arm identification for LLM evaluation, and prediction-powered inference show that informative priors can substantially reduce confidence-sequence width and sampling effort while retaining anytime-valid coverage.

Large Language Models (LLMs) are increasingly used in settings where reliable self-assessment is critical. Assessing model reliability has evolved from using probabilistic correctness estimates to, more recently, eliciting verbalized confidence. Confidence, however, has been shown to be an inconsistent and overoptimistic predictor of model correctness. Drawing on cognitive appraisal theory, a framework from human psychology that decomposes self-evaluation into multiple components, we propose a multidimensional perspective on model self-assessment. We elicit six appraisal-based dimensions of self-assessment, alongside confidence, and evaluate their utility for predicting model failure across 12 LLMs and 38 tasks spanning eight domains. We find that competence-related appraisal dimensions, particularly effort and ability, consistently match or outperform confidence across most settings. Effort additionally yields less overoptimistic estimates that remain stable across model sizes. In contrast, affective dimensions provide marginally predictive signals. Furthermore, the most informative dimension varies systematically with task characteristics: effort is most predictive for reasoning-intensive tasks, while ability and confidence dominate on retrieval-oriented tasks. Broadly, our findings indicate that structured multidimensional self-assessment is a promising approach to improving the reliability and safety of language model deployment across diverse real-world settings.

Large language models are increasingly used as surrogate models for low-data optimization, but their optimizer-facing prediction and its uncertainty remain poorly understood. We study the surrogate belief elicited from an LLM under sparse observations, showing that it depends strongly on prompt text and query protocol. We introduce an uncertainty-alignment criterion that measures whether model uncertainty tracks residual ambiguity among sample-consistent functions. Across controlled inference tasks and Bayesian optimization studies, we find that structural prompts act as effective priors, POINTWISE and JOINT querying induce different beliefs, and sequential evidence leads to non-monotonic, order-sensitive confidence updates. These effects change downstream acquisition decisions and regret, showing that elicitation protocol is part of the LLM surrogate specification, not a formatting detail.

Protein language models are increasingly used to guide experimental and clinical decisions, yet it is often unclear whether a confident prediction reflects recognition of biological evidence or retrieval of a statistical default. We examine this distinction for a near-universal biological rule, that proteins begin with methionine, by tracing the computational pathway through which ESM2-8M produces this prediction. The model does not detect methionine at the masked position. Instead, it retrieves a methionine-favouring signal from a reference representation at the beginning-of-sequence token via a position-specific query assembled across layers, with the final output emerging through competition with context-dependent circuits. To understand how positional information reaches the readout, we introduce a norm-direction decomposition of attention scores within rotary frequency bands. Positional encoding operates through coupled changes in query norm and angular alignment distributed across these bands. On sequences whose true N-terminus is not methionine, where the biological question matters, the model predicts methionine anyway. This is not a correct prediction produced by an unexpected mechanism, but the output of a positional-prior retrieval circuit that matches the statistical average and fails where biology diverges from it. Distinguishing the two requires resolution at the level of individual circuits, frequency bands, and query composition, suggesting that mechanistic verification will be necessary, and challenging, for predictions where the biological stakes are higher. Even for the simplest biological rule, the model's prediction is mediated by a distributed computational circuit rather than direct recognition, suggesting that increasing task complexity will further obscure the relationship between model confidence and underlying biological evidence.

Incorporating specific knowledge into large language models via retrieval-augmented generation (RAG) is a widespread technique that fuels many of today's industry AI applications. A fundamental problem is to assess if the context retrieved by some similarity search provides indeed supporting facts, or instead misguides the generator with irrelevant information. It is critical to associate meaningful confidence measures about the factuality of the retrieval process with the generated answers. We present a new, two-staged approach to predict fact faithfulness of the output of retrieval-augmented generations. First, we employ conformal prediction to select only those retrieved chunks who have a high chance to come from the correct source. This approach in itself can improve answer quality by up to 6% in some of the studied datasets, however, the associated statistical guarantees do not hold generally, since the assumption of sample exchangeability depends on the retriever setup. We present diagnostic metrics to assess whether a setup is suitable. Second, we quantify confidence in the consistency of a generated final answer with a given retrieved context, using an attention-based factuality classifier. This approach can detect inconsistent answers with a chance of up to 77%. Our work helps to establish a novel type of certified RAG systems for a broad range of natural language industry applications.

Confidence calibration has been dominated by the Expected Calibration Error (ECE), a linear metric that counts calibration offset equally regardless of the confidence level at which it occurs. We show that ECE can remain small even under arbitrarily large overconfidence risk, so we propose Calibrated Size Ratio (CSR) instead, an interpretable metric that equals 1 under perfect calibration, from which we derive the risk probability $P_{\mathrm{risk}}$ that quantifies the statistical evidence for overconfidence. We further argue that overconfidence risk assessment must be complemented by a measure of discriminative value: whether the assigned confidences actively distinguish correct from incorrect predictions. We show that confidence-weighted accuracy $\mathrm{cwA}$ is the natural such complement, and that confidence-weighting extends to all standard classification metrics. In particular, we prove that the confidence-weighted AUC (cwAUC) captures the information about calibration while the classical AUC cannot. We validate the proposed indicators on several synthetic confidence distributions under multiple controlled calibration profiles and find that CSR separates risky from non-risky assignments. We also test the metrics on fifteen real datasets, with and without post-hoc calibration, and find that standard methods can yield risky confidence profiles.