Multicalibration and Calibration

We study online multicalibration beyond the worst-case. We give a single, efficient algorithm which dynamically interpolates between benign and worst-case sequences by adaptively refining a dyadic grid of prediction values. Its error is controlled by the number of leaves in the refinement tree. Our analysis recovers the known $\widetilde O(T^{2/3})$ worst-case-optimal rate for online multicalibration, while simultaneously automatically adapting to easier instances: in the marginal stochastic setting it obtains a rate of $\widetilde O(\sqrt T)$, and for piecewise-stationary means with $J$ segments its rate is $\widetilde O(\sqrt{JT})$. More generally, the rate depends on a threshold-complexity measure of the predictable mean process relative to the group family. We show that this dependence is tight up to logarithmic factors.

Multi-camera multi-object tracking (MCMOT) faces significant challenges in maintaining consistent object identities across varying camera perspectives, particularly when precise calibration and extensive annotations are required. In this paper, we present CalibFree, a self-supervised representation learning framework that does not need any calibration or manual labeling for the MCMOT task. By promoting feature separation between view-agnostic and view-specific representations through single-view distillation and cross-view reconstruction, our method adapts to complex, dynamic scenarios with minimal overhead. Experiments on the MMP-MvMHAT dataset show a 3% improvement in overall accuracy and a 7.5% increase in the average F1 score over state-of-the-art approaches, confirming the effectiveness of our calibration-free design. Moreover, on the more diverse MvMHAT dataset, our approach demonstrates superior over-time tracking and strong cross-view performance, highlighting its adaptability to a wide range of camera configurations. Code will be publicly available upon acceptance.

Calibration measures whether a model's predicted confidence aligns with its empirical accuracy, and is central to the reliable deployment of large language models (LLMs) in high-stakes domains such as medicine and law. While much recent work focuses on improving LLM calibration, the equally important question of how to evaluate it in realistic settings remains underdeveloped. Open-ended question answering (QA), the most common deployment setting for modern LLMs, is where existing evaluation methods fall short: logit-based metrics need restricted output formats and internal probabilities; verbalized confidence is self-reported and often overconfident; and sampling-based methods rely on task-specific extraction rules without a clear finite-sample target. We introduce Sem-ECE (Semantic-Sampling Expected Calibration Error), a calibration evaluation framework for open-ended QA that samples answers from the model, groups them into semantic classes, and uses the resulting frequencies as confidence. We study two estimators within this framework: Sem$_1$-ECE, the same-sample self-consistency score, and Sem$_2$-ECE, a held-out variant that separates answer selection from confidence evaluation. We prove both are asymptotically unbiased, and further show that they agree on easy questions but diverge on hard ones with Sem$_2$ achieving strictly smaller calibration error, so their gap also serves as a diagnostic for question difficulty. Experiments on three open-ended QA benchmarks across five leading commercial LLMs match our theoretical predictions and show that Sem-ECE outperforms verbalized confidence and existing sampling-based methods, while complementing logit-based evaluation when internal probabilities are unavailable.

Knowledge Graph Question Answering (KGQA) offers grounded, interpretable reasoning, but existing methods often fail to provide reliable coverage guarantees over retrieved answers. While Conformal Prediction (CP) offers a principled framework for producing prediction sets with statistical guarantees, prior conformal KGQA methods suffer from two critical pitfalls: violated coverage guarantees due to invalid calibration, and weak score discriminability that yields excessively large prediction sets. We propose Conformal Path Reasoning (CPR), a novel trustworthy KGQA framework built on two key innovations. First, query-level conformal calibration over path-level scores preserves exchangeability to ensure valid coverage guarantees. Second, we introduce the Residual Conformal Value Network (RCVNet), a lightweight module trained via PUCT-guided exploration to learn discriminative path-level nonconformity scores. Extensive experiments show that CPR significantly improves the Empirical Coverage Rate by 45% while reducing prediction set size by 52% on average over conformal baselines across benchmark datasets, highlighting its effectiveness for reliable conformal reasoning over knowledge graphs.

We characterize the minimax rate of estimating the second-order calibration error for binary classification, which quantifies whether a higher-order predictor's epistemic-uncertainty estimate matches the conditional variance of the label probability on its level sets. Our key observation is that the sech perturbation kernel, previously used only to enforce smoothness of calibration functions, in fact makes them analytic in a strip of half-width $hπ/2$. Polynomial regression then estimates the calibration error at rate $\tilde{O}(1/\sqrt{n})$, with explicit constants, a qualitative improvement over the $O(n^{-1/4})$ rate achievable by bucketing or kernel smoothing. A matching $Ω(1/\sqrt{n})$ lower bound establishes minimax optimality up to logarithmic factors. As a corollary, we give the first finite-sample guarantee for second-order Platt scaling, yielding a post-hoc procedure that recalibrates both the mean prediction and the epistemic-variance estimate of any higher-order predictor. Along the way, we provide a bucket-free definition of second-order calibration and relate it quantitatively to the bucketed formulation of Ahdritz et al. [2025]. Our experiments confirm the predicted rate and the quality of the recalibrated uncertainties.

This paper proposes a novel framework for implicit multi-camera system calibration utilizing Gaussian Process (GP) regression. Conventional explicit calibration methods are constrained by rigid mathematical models and struggle with complex, non-linear distortions from unconventional optics, while existing neural network-based implicit approaches are typically data-hungry and lack inherent uncertainty quantification (UQ). Our GP-based model directly learns the complex, non-linear mapping from 2D image coordinates across all cameras to a 3D world coordinate, completely bypassing time-consuming estimation of explicit intrinsic and extrinsic parameters. Moreover, the inherent UQ is critical for transforming a simple 3D point prediction into a verifiable 3D measurement, complete with statistically-sound confidence bounds. To further enhance data efficiency and practical deployment, we integrate Active Learning (AL), which intelligently leverages the GP's predictive uncertainty to strategically guide the acquisition of new calibration data. This approach results in a robust, data-efficient, and reliable calibration solution, proving particularly effective in practical scenarios where collecting extensive calibration data is a dominant constraint. Our experiments show that the uncertainty for the 3D predictions is higher closer to the cameras. The data points in $uv$-coordinate space are more sparse in that region, even though they are not in 3D space. This work is relevant for anyone who is tasked with the calibration of complex multi-camera systems.

AI agents that plan, retain memory across sessions, invoke external tools and act with partial autonomy are transforming human--AI collaboration. Research on affective computing, simulated empathy in large language models, trust in automation and AI safety has illuminated important design principles, yet these literatures remain fragmented. No integrated account explains how affective cues operate within agentic collaboration -- settings in which humans delegate, monitor and correct consequential tasks. This Review synthesises computational and interactional mechanisms of affective dynamics: the processes through which affective cues, emotion-like behaviour and perceived agent affect shape trust calibration, delegation decisions, error correction, dependence and governance. We trace how model-generated affective signals enter interaction loops that govern reliance, repair and oversight, and propose a framework that treats affect not as an internal property of AI but as a coordination layer through which humans and agents negotiate capability, uncertainty and responsibility. The framework provides a foundation for calibrated measurement, purposeful design and informed governance.

The Translation Tax is often treated as a scalar: translated benchmarks are assumed to inflate scores by preserving English-source cues. We audit this claim in an English-to-Chinese setting. Three proxy estimators disagree: back-translation gaps are small and parser-fragile; cue-score calibration does not predict item-level gains; and a six-model native-control comparison shows model-family rather than uniform benchmark effects. We add a same-item LLM-naturalization stress test that holds answer, options, and content fixed while rewriting Chinese surface form. After correcting a prompt-construction bug, this contrast no longer supports a model-family interaction, but it preserves a residue dose-response: high-residue items benefit while low-residue items do not. The result is not a single Translation Tax, but a set of estimator- and item-dependent validity risks. We release per-cell evidence, the naturalization protocol, human QC, and a reporting checklist for translated multilingual benchmark papers.

This methods article presents a reproducible calibration workflow for prompt-based large language models (LLMs) in structured evidence-synthesis tasks. The method separates the rules that define the scientific task from the mutable prompt harness that frames and applies them. It optimises that harness against labelled or reference examples and an explicit task metric, then preserves the calibrated workflow as an inspectable artefact with its specification, metric, settings, and evaluation traces. The example code instantiates the protocol with DSPy and GEPA tools, but the underlying logic can transfer to other prompt-optimisation frameworks that support structured task definitions, metric-guided search, and artefact reuse. Title and abstract screening is the worked validation case because it provides labelled benchmark data and clear evaluation metrics. The demonstrated workflow uses a smaller student LLM for performing the scientific task execution and a larger reflection LLM to steer the prompt optimisation process during calibration. This work shows compilation, artefact round-tripping, and how optimisation budget affects a smaller student model.

Post-training quantization (PTQ) is an effective approach for deploying large language models (LLMs) under memory and latency constraints. Most existing PTQ methods determine quantization parameters by minimizing a layer-wise reconstruction error on a predetermined calibration dataset, typically optimized via either scale search or Gram-based methods. However, from the perspective of generalization risk, existing PTQ calibration objectives based solely on empirical reconstruction error over limited or unrepresentative calibration data may move the quantized weights away from the original floating-point weights, potentially degrading downstream performance. To address this issue, we propose \emph{Regularized Quantization Calibration} (RQC), a unified framework that augments standard PTQ objectives with a regularizer that explicitly controls weight deviation from the original weights. We further generalize this framework to incorporate a saliency-aware regularizer, resulting in \emph{Saliency-Aware Regularized Quantization Calibration} (SARQC). The proposed regularization encourages quantized weights to remain close to the original weights during calibration, leading to improved generalization at inference time. SARQC integrates seamlessly into existing PTQ pipelines and enhances both scale-search-based and Gram-based methods under a unified formulation. Extensive experiments on dense and Mixture-of-Experts LLMs demonstrate consistent improvements in perplexity and zero-shot accuracy, without introducing additional inference overhead.

Label Distribution Learning (LDL) models supervision as an instance-wise probability distribution, enabling fine-grained learning under inherent ambiguity, but its success relies on high-fidelity label distributions that are costly to obtain and thus often noisy. Motivated by privacy-sensitive applications, we study Federated Label Distribution Learning (Fed-LDL), where data isolation further induces heterogeneous annotation quality across clients, making local updates unevenly reliable and breaking sample-size-based aggregation (e.g., FedAvg). To address this trust dilemma, we propose FedQual, a quality-aware Fed-LDL framework with two coupled mechanisms: (i) quality-adaptive client training guided by a global semantic anchor that calibrates low-quality clients while preserving high-quality autonomy, and (ii) reliability-aware server aggregation that reweights client contributions by effective reliable information rather than raw sample size. To enable rigorous evaluation, we construct four new Fed-LDL benchmarks (FER-LDL, FI-LDL, PIPAL-LDL, and KADID-LDL) with controlled annotation quality disparity. We further provide a theoretical guarantee showing that under heterogeneous supervision quality, client-specific calibration is strictly better than any uniform calibration. Extensive experiments on the proposed benchmarks demonstrate the effectiveness of FedQual.

This article proposes a general optimization framework for solving hand-eye calibration problem. Unlike traditional methods, an iterative algorithm based on Lie algebra that achieves approximately global optimal solutions is developed. During the optimization process, the method strictly preserves the structural constraints of the calibration parameters and enables synchronized updates between calibration parameters. Recognizing that data used in real-word hand-eye calibration often contain uncertainty, especially in over-loading and large workspace industrial robot scenarios, which can significantly degrade accuracy, and accurately modeling such uncertainty is inherently difficult, this article avoids explicit uncertainty modeling. Instead, an uncertainty metric to evaluate the relative uncertainty between data sources is introduced and used to dynamically refine the iterative process. To further enhance convergence efficiency, an effective initial solution generation method that improves overall stability and accuracy is designed. Numerical simulations and real-world experiments validate the effectiveness of the proposed approach, and in synthetic datasets, the proposed approach improves the estimation accuracy by at least 67% under high-uncertainty conditions compared with the existing methods.

Emergency department triage assigns patients an acuity score that determines treatment priority, and clinical evidence documents persistent gender disparities in human acuity assessment. As hospitals pilot large language models (LLMs) as triage decision support, a critical question is whether these models reproduce or mitigate known biases. We present EQUITRIAGE, a fairness audit of LLM-based ESI assignment evaluating five models (Gemini-3-Flash, Nemotron-3-Super, DeepSeek-V3.1, Mistral-Small-3.2, GPT-4.1-Nano) across 374,275 evaluations on 18,714 MIMIC-IV-ED vignettes under four prompt strategies. Of 9,368 originals, 9,346 are paired with a gender-swapped counterfactual. All five models produced flip rates above a pre-registered 5% threshold (9.9% to 43.8%). Two showed directional female undertriage (DeepSeek F/M 2.15:1, Gemini 1.34:1); two were near-parity; one had high sensitivity with weak male-direction asymmetry. DeepSeek's directional bias coexisted with a low outcome-linked calibration gap (0.013 against MIMIC-IV admission), a Chouldechova-style dissociation between within-group calibration and between-pair counterfactual invariance. Demographic blinding reduced Gemini's flip rate to 0.5%; an age-preserving blind variant left DeepSeek with residual F/M 1.25, implicating age as a residual channel. Chain-of-thought prompting degraded accuracy for all five models. A two-model ablation reveals opposite underlying mechanisms for the same directional phenotype: in Gemini the signal is emergent in the combined name+gender swap, while in DeepSeek the gender token alone carries it. EQUITRIAGE shows that group parity, counterfactual invariance, and gender calibration are distinct fairness properties, that intervention effectiveness is model-dependent, and that per-model counterfactual auditing should precede clinical deployment.

Multimodal learning often grapples with the challenge of low-quality data, which predominantly manifests as two facets: modality imbalance and noisy corruption. While these issues are often studied in isolation, we argue that they share a common root in the predictive uncertainty towards the reliability of individual modalities and instances during learning. In this paper, we propose a unified framework, termed Conformal Predictive Self-Calibration (CPSC), which leverages conformal prediction to equip the model with the ability to perform self-guided calibration on-the-fly. The core of our proposed CPSC lies in a novel self-calibrating training loop that seamlessly integrates two key modules: (1) Representation Self-Calibration, which decomposes unimodal features into components, and selectively fuses the most robust ones identified by a conformal predictor to enhance feature resilience. (2) Gradient Self-Calibration, which recalibrates the gradient flow during backpropagation based on instance-wise reliability scores, steering the optimization towards more trustworthy directions. Furthermore, we also devise a self-update strategy for the conformal predictor to ensure the entire system co-evolves consistently throughout the training process. Extensive experiments on six benchmark datasets under both imbalanced and noisy settings demonstrate that our CPSC framework consistently outperforms existing state-of-the-art methods. Our code is available at https://github.com/XunCHN/CPSC.

Open-vocabulary object detection often fails under distribution shifts, as it can be misled by spurious correlations between non-causal visual attributes (e.g., brightness, texture) and object categories. Existing test-time adaptation (TTA) methods either depend on costly online optimization or perform global calibration, overlooking the attribute-specific nature of these failures. To address this, we propose FACTOR (counterFACtual training-free Test-time adaptation for Open-vocabulaRy object detection), a lightweight framework grounded in counterfactual reasoning. By perturbing test images along non-causal attributes and comparing region-level predictions between original and counterfactual views, FACTOR quantifies attribute sensitivity, semantic relevance, and prediction variation to selectively suppress attribute-dependent predictions-without parameter updates. Experiments on PASCAL-C, COCO-C, and FoggyCityscapes show that FACTOR consistently outperforms prior TTA methods, demonstrating that explicit counterfactual reasoning effectively improves robustness under distribution shifts.

Hyperledger Fabric performance depends on many interacting configuration parameters, making manual tuning difficult. We study automated throughput tuning by treating benchmarking as a noisy black-box optimization problem and applying Bayesian optimization (BO) with dimensionality reduction (DR). We implement an end-to-end Caliper-in-the-loop pipeline that deploys candidate configurations, benchmarks them, and updates the optimizer from observed throughput. The search space, derived from Fabric configuration files, has 317 dimensions. In a cloud testbed, we evaluate 16 BO+DR variants and a random-search baseline. The best method, DYCORS-PCA, achieves a 12% TPS improvement relative to the first evaluated configuration, while MPI-REMBO achieves 9%. These results suggest that BO with DR is a practical approach for high-dimensional Hyperledger Fabric tuning, while also highlighting the role of measurement noise in interpreting gains.

Models that are indistinguishable on in-distribution data can behave very differently under distribution shift. We introduce Perturb-and-Correct (P&C), a post-hoc method for constructing epistemically diverse predictors from a single pretrained network. P&C applies random hidden layer perturbations with a least-squares correction in the subsequent affine layer, producing predictors that agree on calibration data while remaining free to disagree away from it. We analyze this mechanism through the post-correction residual and its first-order sensitivity: the residual is controlled near the calibration distribution by a leverage term, while corrected sensitivity grows as inputs deviate from the calibration geometry. Empirically, P&C achieves a strong ID/OOD tradeoff across MuJoCo dynamics prediction and CIFAR-10 OOD detection, matching or outperforming standard post-hoc baselines while requiring only a single pretrained model. Our findings highlight the potential in further exploiting overparameterization as a strength of deep learning models.

The building energy community lacks a foundational thermal model, i.e., a single pretrained model capable of generalizing across diverse buildings, climates, and control strategies without building-specific calibration. Achieving this vision requires architectural principles that capture universal thermal dynamics rather than memorizing building-specific patterns. We take a step toward this goal by presenting a physics-informed transformer architecture that embeds domain knowledge, e.g., derivative enrichment and Euler-based numerical integration, into a decoder-only framework. We incorporate static building features extracted from simulation models and employ Rotary Position Embedding attention to capture temporal dependencies. Evaluated on the CityLearn dataset spanning 247 residential buildings across three climate zones, our model achieves one-step prediction accuracy (RMSE of 0.30°C in Texas, 0.29°C in Vermont) while outperforming both traditional baselines and fine-tuned Time-Series Foundation Models. We also demonstrate zero-shot transferability: models trained on as few as two buildings generalize to unseen buildings and climate zones without fine-tuning. Despite the limitation of simulated residential buildings, our results establish physics-informed architectural principles as a promising foundation for universal building thermal models.

Large Vision-Language Models (LVLMs) often suffer from hallucinations, generating descriptions that include visual details absent from the input image. Recent preference alignment methods typically rely on supervision distilled from stronger models such as GPT. However, this offline paradigm introduces a Supervision-Perception Mismatch: the student model is forced to align with fine-grained details beyond its perceptual capacity, learning to guess rather than to see. To obtain reliable self-supervision for online learning, we identify a Generative-Discriminative Gap within LVLMs, where models exhibit higher accuracy on discriminative verification than open-ended generation. Leveraging this capability, we propose \textbf{O}nline \textbf{S}elf-\textbf{CA}lib\textbf{R}ation (OSCAR), a framework that integrates Monte Carlo Tree Search with a Dual-Granularity Reward Mechanism to construct preference data and iteratively refines the model via Direct Preference Optimization. Extensive experiments demonstrate that OSCAR achieves state-of-the-art performance on hallucination benchmarks while improving general multimodal capabilities.

Test-time prompt tuning (TPT) has emerged as a promising technique for enhancing the adaptability of vision-language models by optimizing textual prompts using unlabeled test data. However, prior studies have observed that TPT often produces poorly calibrated models, raising concerns about the reliability of their predictions. Recent works address this issue by incorporating additional regularization terms that constrain model outputs, which improve calibration but often degrade performance. In this work, we reveal that these regularization strategies implicitly encourage optimization toward flatter minima, and that the sharpness of the loss landscape around adapted prompts is a key factor governing calibration quality. Motivated by this observation, we introduce Flatness-aware Prompt Pretraining (FPP), a simple yet effective pretraining framework for TPT that initializes prompts within flatter regions of the loss landscape prior to adaptation. We show that simply replacing the initialization in existing TPT pipelines--without modifying any other components--is sufficient to improve both calibration and performance. Notably, FPP requires no labeled data and incurs no additional computational costs during test-time tuning, making it highly practical for real-world deployment. The code is available at: https://github.com/YonseiML/fpp.

Safety-critical prediction systems, such as autonomous vehicles, weather forecasters, and medical monitors, commonly rely on probabilistic forecasters. These forecasters make predictions about possible future outcomes, and their quality and robustness needs to be validated and certified. Often, only accuracy -- the mean of the predictions -- is evaluated against true outcomes. However, for safety-critical scenarios and decision making under uncertainty, the full distributional properties of the forecasts should be checked: do the observed prediction errors actually follow the forecasted probability distributions? To this end, we introduce a framework for calibration checks: statistical tests that validate distributional properties of forecasts when measured over many samples. In order to support ease-of-use in real-world operations, these checks produce a single accept/reject decision for data collected from a forecaster. This contrasts typical calibration calculations which produce one or multiple continuous calibration scores and require expertise to implement in a validation workflow. We further support operationalization by introducing modifications to calibration testing that (a) reject only overconfident predictions, allowing for pessimistic or cautious predictions in safety-critical settings, and (b) tolerate small, operationally acceptable deviations even for large numbers of validation samples. We organize the calibration checking process into a modular pipeline comprising four steps: (i) the data model, (ii) the chosen metric, (iii) the hypothesis formulation, and (iv) the testing procedure. Each step consists of independently swappable components, thereby supporting a large variety of possible use-cases and trade-offs. We demonstrate the applicability of the framework on two complementary example problems, weather forecasting and robot pose estimation.

Quantum computing calibration depends on interpreting experimental data, and calibration plots provide the most universal human-readable representation for this task, yet no systematic evaluation exists of how well vision-language models (VLMs) interpret them. We introduce QCalEval, the first VLM benchmark for quantum calibration plots: 243 samples across 87 scenario types from 22 experiment families, spanning superconducting qubits and neutral atoms, evaluated on six question types in both zero-shot and in-context learning settings. The best general-purpose zero-shot model reaches a mean score of 72.3, and many open-weight models degrade under multi-image in-context learning, whereas frontier closed models improve substantially. A supervised fine-tuning ablation at the 9-billion-parameter scale shows that SFT improves zero-shot performance but cannot close the multimodal in-context learning gap. As a reference case study, we release NVIDIA Ising Calibration 1, an open-weight model based on Qwen3.5-35B-A3B that reaches 74.7 zero-shot average score.

Continual learning for large language models is typically evaluated through accuracy retention under sequential fine-tuning. We argue that this perspective is incomplete, because uncertainty reliability can degrade earlier and more sharply than top-1 performance. We study this empirically by measuring conformal coverage and calibration error on sequentially fine-tuned models across three model families and eight task sequences drawn primarily from classification and multiple-choice benchmarks. Across the classification-style settings we study, coverage loss exceeds accuracy loss by a factor of roughly $3.4\times \pm 0.5\times$ on average across seeds; in the most pronounced case, coverage drops from $0.92$ to $0.61$, while accuracy remains within three points of baseline. Standard continual-learning methods that preserve accuracy do not automatically preserve coverage, and naive calibration baselines recover only part of the gap. We propose calibration replay, a lightweight post-hoc procedure that maintains a task-specific held-out buffer and refits a task-specific conformal threshold under the current model after each update. It adds no training-time gradient cost, uses less than one percent of the memory of ordinary experience replay, and typically restores coverage to within two points of nominal at buffer size $m = 200$. We accompany the empirical study with a drift decomposition, a finite-sample recovery theorem showing exact conformal validity under exchangeability, and a mixture-validity proposition explaining why pooled thresholds do not suffice. Our guarantees are stated for classification-style tasks with task-specific buffers; extensions to open-ended generation are exploratory.

We study the minimax sample complexity of multicalibration in the batch setting. A learner observes $n$ i.i.d. samples from an unknown distribution and must output a (possibly randomized) predictor whose population multicalibration error, measured by Expected Calibration Error (ECE), is at most $\varepsilon$ with respect to a given family of groups. For every fixed $κ> 0$, in the regime $|G|\le \varepsilon^{-κ}$, we prove that $\widetildeΘ(\varepsilon^{-3})$ samples are necessary and sufficient, up to polylogarithmic factors. The lower bound holds even for randomized predictors, and the upper bound is realized by a randomized predictor obtained via an online-to-batch reduction. This separates the sample complexity of multicalibration from that of marginal calibration, which scales as $\widetildeΘ(\varepsilon^{-2})$, and shows that mean-ECE multicalibration is as difficult in the batch setting as it is in the online setting, in contrast to marginal calibration which is strictly more difficult in the online setting. In contrast we observe that for $κ= 0$, the sample complexity of multicalibration remains $\widetildeΘ(\varepsilon^{-2})$ exhibiting a sharp threshold phenomenon. More generally, we establish matching upper and lower bounds, up to polylogarithmic factors, for a weighted $L_p$ multicalibration metric for all $1 \le p \le 2$, with optimal exponent $3/p$. We also extend the lower-bound template to a regular class of elicitable properties, and combine it with the online upper bounds of Hu et al. (2025) to obtain matching bounds for calibrating properties including expectiles and bounded-density quantiles.

We study semisupervised mean estimation with a small labeled sample, a large unlabeled sample, and a black-box prediction model whose output may be miscalibrated. A standard approach in this setting is augmented inverse-probability weighting (AIPW) [Robins et al., 1994], which protects against prediction-model misspecification but can be inefficient when the prediction score is poorly aligned with the outcome scale. We introduce Calibrated Prediction-Powered Inference, which post-hoc calibrates the prediction score on the labeled sample before using it for semisupervised estimation. This simple step requires no retraining and can improve the original score both as a predictor of the outcome and as a regression adjustment for semisupervised inference. We study both linear and isotonic calibration. For isotonic calibration, we establish first-order optimality guarantees: isotonic post-processing can improve predictive accuracy and estimator efficiency relative to the original score and simpler post-processing rules, while no further post-processing of the fitted isotonic score yields additional first-order gains. For linear calibration, we show first-order equivalence to PPI++. We also clarify the relationship among existing estimators, showing that the original PPI estimator is a special case of AIPW and can be inefficient when the prediction model is accurate, while PPI++ is AIPW with empirical efficiency maximization [Rubin et al., 2008]. In simulations and real-data experiments, our calibrated estimators often outperform PPI and are competitive with, or outperform, AIPW and PPI++. We provide an accompanying Python package, ppi_aipw, at https://larsvanderlaan.github.io/ppi-aipw/.

We give a Gordon-Greenwald-Marks (GGM) style black-box reduction from online learning to online multicalibration. Concretely, we show that to achieve high-dimensional multicalibration with respect to a class of functions H, it suffices to combine any no-regret learner over H with an expected variational inequality (EVI) solver. We also prove a converse statement showing that efficient multicalibration implies efficient EVI solving, highlighting how EVIs in multicalibration mirror the role of fixed points in the GGM result for $Φ$-regret. This first set of results resolves the main open question in Garg, Jung, Reingold, and Roth (SODA '24), showing that oracle-efficient online multicalibration with $\sqrt{T}$-type guarantees is possible in full generality. Furthermore, our GGM-style reduction unifies the analyses of existing online multicalibration algorithms, enables new algorithms for challenging environments with delayed observations or censored outcomes, and yields the first efficient black-box reduction between online learning and multiclass omniprediction. Our second main result is a fine-grained reduction from high-dimensional online multicalibration to (contextual) $Φ$-regret minimization. Together with our first result, this establishes a new route from external regret to Phi-regret that bypasses sophisticated fixed-point or semi-separation machinery, dramatically simplifies a result of Daskalakis, Farina, Fishelson, Pipis, and Schneider (STOC '25) while improving rates, and yields new algorithms that are robust to richer deviation classes, such as those belonging to any reproducing kernel Hilbert space.

Transformer-based scientific foundation models are increasingly deployed in high-stakes settings, but current architectures give deterministic outputs and provide limited support for calibrated predictive uncertainty. We propose Stochastic Attention, a sample average lightweight inference-time modification that randomizes attention by replacing softmax weights with normalized multinomial samples controlled by a single concentration parameter, and produces predictive ensembles without retraining. To set this parameter, we introduce a calibration objective that matches the stochastic attention output with the target, yielding an efficient univariate post-hoc tuning problem. We evaluate this mechanism on scientific foundation models for weather and time-series forecasting, as well as several regression tasks. Across benchmarks against uncertainty-aware baselines, we find that Sample Average Stochastic Attention achieves the strongest native calibration and the sharpest prediction intervals at comparable calibration, with adaptation costs nearly three orders of magnitude lower than the next-best baseline.

Knowledge distillation (KD) transfers capabilities from large language models (LLMs) to smaller students, yet it can fail unpredictably and also underpins model leakage risks. Our analysis revealed several distillation traps: tail noise, off-policy instability, and, most fundamentally, the teacher-student gap, that distort training signals. These traps manifest as overconfident hallucinations, self-correction collapse, and local decoding degradation, causing distillation to fail. Motivated by these findings, we propose a post-hoc calibration method that, to the best of our knowledge, for the first time enables control over a teacher's distillability via reinforcement fine-tuning (RFT). Our objective combines task utility, KL anchor, and across-tokenizer calibration reward. This makes distillability a practical safety lever for foundation models, connecting robust teacher-student transfer with deployment-aware model protection. Experiments across math, knowledge QA, and instruction-following tasks show that students distilled from distillable calibrated teachers outperform SFT and KD baselines, while undistillable calibrated teachers retain their task performance but cause distilled students to collapse, offering a practical knob for both better KD and model IP protection.

Post-training quantization (PTQ) is essential for deploying large diffusion transformers on resource-constrained hardware, but aggressive 4-bit quantization significantly degrades generative performance. Low-rank approximation methods have emerged as a promising solution by appending auxiliary linear branches to restore performance. However, current state-of-the-art approaches assume these branches must retain high precision (W16A16) and rely on heavy, data-dependent calibration for initialization. We challenge both limitations with LoRaQ (Low-Rank Approximated Quantization), a simple, data-free calibration approach that optimizes quantization error compensation. By overcoming the need for high-precision branches, LoRaQ enables the first fully sub-16 bit pipeline, allowing the low-rank branch itself to be quantized. We demonstrate that, at equal memory overhead, LoRaQ outperforms the state-of-the-art methods in their native implementations on Pixart-$Σ$ and SANA. We also analyze mixed-precision configurations, showing that setups such as W8A8, W6A6, and W4A8 for the low-rank branch, alongside a W4 main layer, yield superior results while maintaining a fully quantized architecture compatible with modern mixed-precision hardware.

We revisit multi-agent delegation under a stronger and more realistic assumption: an agent's capability is not fixed at the skill level, but depends on task context. A coding agent may excel at short standalone edits yet fail on long-horizon debugging; a planner may perform well on shallow tasks yet degrade on chained dependencies. Static skill-level capability profiles therefore average over heterogeneous situations and can induce systematic misdelegation. We propose CADMAS-CTX, a framework for contextual capability calibration. For each agent, skill, and coarse context bucket, CADMAS-CTX maintains a Beta posterior that captures stable experience in that part of the task space. Delegation is then made by a risk-aware score that combines the posterior mean with an uncertainty penalty, so that agents delegate only when a peer appears better and that assessment is sufficiently well supported by evidence. This paper makes three contributions. First, a hierarchical contextual capability profile replaces static skill-level confidence with context-conditioned posteriors. Second, based on contextual bandit theory, we formally prove context-aware routing achieves lower cumulative regret than static routing under sufficient context heterogeneity, formalizing the bias-variance tradeoff. Third, we empirically validate our method on GAIA and SWE-bench benchmarks. On GAIA with GPT-4o agents, CADMAS-CTX achieves 0.442 accuracy, outperforming static baseline 0.381 and AutoGen 0.354 with non-overlapping 95% confidence intervals. On SWE-bench Lite, it improves resolve rate from 22.3% to 31.4%. Ablations show the uncertainty penalty improves robustness against context tagging noise. Our results demonstrate contextual calibration and risk-aware delegation significantly improve multi-agent teamwork compared with static global skill assignments.

With the rapid growth of video data, Composed Video Retrieval (CVR) has emerged as a novel paradigm in video retrieval and is receiving increasing attention from researchers. Unlike unimodal video retrieval methods, the CVR task takes a multi-modal query consisting of a reference video and a piece of modification text as input. The modification text conveys the user's intended alterations to the reference video. Based on this input, the model aims to retrieve the most relevant target video. In the CVR task, there exists a substantial discrepancy in information density between video and text modalities. Traditional composition methods tend to bias the composed feature toward the reference video, which leads to suboptimal retrieval performance. This limitation is significant due to the presence of three core challenges: (1) modal contribution entanglement, (2) explicit optimization of composed features, and (3) retrieval uncertainty. To address these challenges, we propose the evidence-dRivRn dual-sTream diRectionAl anChor calibration networK (ReTrack). ReTrack is the first CVR framework that improves multi-modal query understanding by calibrating directional bias in composed features. It consists of three key modules: Semantic Contribution Disentanglement, Composition Geometry Calibration, and Reliable Evidence-driven Alignment. Specifically, ReTrack estimates the semantic contribution of each modality to calibrate the directional bias of the composed feature. It then uses the calibrated directional anchors to compute bidirectional evidence that drives reliable composed-to-target similarity estimation. Moreover, ReTrack exhibits strong generalization to the Composed Image Retrieval (CIR) task, achieving SOTA performance across three benchmark datasets in both CVR and CIR scenarios. Codes are available at https://github.com/Lee-zixu/ReTrack

Deploying Vision-Language Models (VLMs) under aggressive low-bit inference remains challenging because inference cost is dominated by the long visual-token prefix during prefill and the growing KV cache during autoregressive decoding. Token pruning and low-bit quantization are complementary for reducing these costs, yet naive stage-wise combinations are often brittle due to a mismatch between quantization calibration and pruning execution. We present a collaborative quantization-and-pruning framework that unifies low-bit inference and deterministic visual-token pruning in a single deployable pipeline. The framework introduces the \textbf{Q}uantization \textbf{U}nified \textbf{O}ffline \textbf{T}oken \textbf{A}llocator (\textbf{QUOTA}), which converts low-bit calibration signals into a layer-wise token allocation schedule and materializes it as a pruning recipe. Token importance is evaluated under deployed W4A4 operators with a quantized KV cache by combining activation magnitude, attention cues, and an explicit low-bit risk signal, enabling consistent budgeted top-$k$ selection. Experiments on standard VLM benchmarks show improved robustness over stage-wise baselines under the same low-bit regime, achieving 95.65% average retention while retaining only 30% of visual tokens, compared with about 94.3% retention for representative stage-wise combinations. The code will be released.

Recent advances in summary evaluation are based on model-based metrics to assess quality dimensions, such as completeness, conciseness, and faithfulness. However, these methods often require large language models, and predicted scores are frequently miscalibrated, limiting their reliability. Moreover, evaluating the average quality across different summaries for a single document typically requires access to multiple reference summaries. Here, we propose a general framework that generates individual and average proxy scores without relying on reference summaries, human annotations, or expensive model-based metrics. We also propose group isotonic regression binning (GIRB), a calibration method that adjusts the raw predictions to better align with ground-truth evaluation metrics. While we focus on continuous-value scenarios, such as summarization, the method is applicable to discrete-value tasks, such as question answering. Experiments on seven datasets demonstrate that our approach consistently outperforms existing baselines.

Merging separately trained LoRA adapters is a practical alternative to joint multi-task training, but it often hurts performance. Existing methods usually treat the LoRA update $ΔW = BA$ as a single object and do not distinguish the two LoRA matrices. We show that the main source of LoRA merge interference comes from the output-side matrix $B$. Across tasks, $B$ repeatedly uses a small set of shared directions, while $A$ remains much more task-specific. As a result, the merged adapter overemphasizes these shared directions, and task-specific information is lost. We propose Pico (Pre-merge interference calibration in output-space), a data-free method that calibrates $B$ before merge by downscaling over-shared directions and then rescaling the merged update. Pico plugs directly into existing merging methods such as Task Arithmetic, TIES, and TSV-M. Across eight different benchmarks from math, coding, finance, and medical domains, Pico improves average accuracy by 3.4-8.3 points over the corresponding base method and achieves the best overall average performance. Pico also enables merged adapters to outperform the LoRA trained with all task data. These results show that LoRA merging works better when the two LoRA matrices are treated separately.

Hand-eye calibration through visual localization is a critical capability for robotic manipulation in open-world environments. However, most deep learning-based calibration models suffer from catastrophic forgetting when adapting into unseen data amongst open-world scene changes, while simple rehearsal-based continual learning strategy cannot well mitigate this issue. To overcome this challenge, we propose a continual hand-eye calibration framework, enabling robots to adapt to sequentially encountered open-world manipulation scenes through spatially replay strategy and structure-preserving distillation. Specifically, a Spatial-Aware Replay Strategy (SARS) constructs a geometrically uniform replay buffer that ensures comprehensive coverage of each scene pose space, replacing redundant adjacent frames with maximally informative viewpoints. Meanwhile, a Structure-Preserving Dual Distillation (SPDD) is proposed to decompose localization knowledge into coarse scene layout and fine pose precision, and distills them separately to alleviate both types of forgetting during continual adaptation. As a new manipulation scene arrives, SARS provides geometrically representative replay samples from all prior scenes, and SPDD applies structured distillation on these samples to retain previously learned knowledge. After training on the new scene, SARS incorporates selected samples from the new scene into the replay buffer for future rehearsal, allowing the model to continuously accumulate multi-scene calibration capability. Experiments on multiple public datasets show significant anti scene forgetting performance, maintaining accuracy on past scenes while preserving adaptation to new scenes, confirming the effectiveness of the framework.

Significant progress has been made in detecting synthetic images, however most existing approaches operate on a single image instance and overlook a key characteristic of real-world dissemination: as viral images circulate on the web, multiple near-duplicate versions appear and lose quality due to repeated operations like recompression, resizing and cropping. As a consequence, the same image may yield inconsistent forensic predictions based on which version has been analyzed. In this work, to address this issue we propose QuAD (Quality-Aware calibration with near-Duplicates) a novel framework that makes decisions based on all available near-duplicates of the same image. Given a query, we retrieve its online near-duplicates and feed them to a detector: the resulting scores are then aggregated based on the estimated quality of the corresponding instance. By doing so, we take advantage of all pieces of information while accounting for the reduced reliability of images impaired by multiple processing steps. To support large-scale evaluation, we introduce two datasets: AncesTree, an in-lab dataset of 136k images organized in stochastic degradation trees that simulate online reposting dynamics, and ReWIND, a real-world dataset of nearly 10k near-duplicate images collected from viral web content. Experiments on several state-of-the-art detectors show that our quality-aware fusion improves their performance consistently, with an average gain of around 8% in terms of balanced accuracy compared to plain average. Our results highlight the importance of jointly processing all the images available online to achieve reliable detection of AI-generated content in real-world applications. Code and data are publicly available at https://grip-unina.github.io/QuAD/

Edge AI model deployment is a multi-stage engineering process involving model conversion, operator compatibility handling, quantization calibration, runtime integration, and accuracy validation. In practice, this workflow is long, failure-prone, and heavily dependent on deployment expertise, particularly when targeting hardware-specific inference runtimes. This technical report presents AIPC (AI Porting Conversion), an AI agent-driven approach for constrained automation of AI model deployment. AIPC decomposes deployment into standardized, verifiable stages and injects deployment-domain knowledge into agent execution through Agent Skills, helper scripts, and a stage-wise validation loop. This design reduces both the expertise barrier and the engineering time required for hardware deployment. Using Qualcomm AI Runtime (QAIRT) as the primary scenario, this report examines automated deployment across representative vision, multimodal, and speech models. In the cases covered here, AIPC can complete deployment from PyTorch to runnable QNN/SNPE inference within 7-20 minutes for structurally regular vision models, with indicative API costs roughly in the range of USD 0.7-10. For more complex models involving less-supported operators, dynamic shapes, or autoregressive decoding structures, fully automated deployment may still require further advances, but AIPC already provides practical support for execution, failure localization, and bounded repair.

Selective conformal prediction can yield substantially tighter uncertainty sets when we can identify calibration examples that are exchangeable with the test example. In interventional settings, such as perturbation experiments in genomics, exchangeability often holds only within subsets of interventions that leave a target variable "unaffected" (e.g., non-descendants of an intervened node in a causal graph). We study the practical regime where this invariance structure is unknown and must be estimated from data. Our main result quantifies how coverage degrades when the estimated safe calibration set accidentally includes interventions that affect the target, and gives a conservative correction when an upper bound on this error is available. Rather than learning a full causal graph, we learn only the intervention-target relationships needed to choose calibration interventions. We give algorithms for this partial learning task and evaluate them on synthetic structural equation models and Replogle K562 CRISPR-interference data, where the experiments illustrate synthetic gains from selective calibration and finite-sample tradeoffs on real perturbation screens.

Flexible endoscopic continuum manipulators offer high dexterity and access to complex anatomy, but nonlinear hysteresis limits feedforward control accuracy. Closed-loop control can compensate for these errors but requires accurate six-degree-of-freedom (6D) end-effector pose feedback. We present MarUco, a markerless 6D pose estimation framework for closed-loop control using only stereo vision during operation. A photorealistic pseudo-rigid-body simulation pipeline generates large-scale annotated training data without manual labeling. A multifeature fusion network integrates masks, keypoints, heatmaps, and bounding boxes from both stereo views to estimate an initial pose, followed by a learned single-pass render-and-compare module that refines the pose without iterative optimization. Kinematics-free hand-eye calibration estimates camera-to-robot-base extrinsics, and self-supervised adaptation uses unlabeled real stereo pairs to mitigate sim-to-real pose error. Across 1,000 real samples, MarUco achieves translation and rotation errors of 0.78 $\pm$ 0.50 mm and 3.07 $\pm$ 1.25°, respectively, with a total processing time of 52.4 ms per stereo pair. In closed-loop experiments over eight reference paths, MarUco achieves a mean terminal translation error of 1.8 mm, an 88% reduction relative to uncompensated open-loop control. To the best of our knowledge, this is the first markerless position-based visual servoing framework for continuum manipulators.

Running the same LLM agent on identical inputs yields 2.3-4.2 distinct action sequences per 10 runs; this behavioral variance constitutes a training-free, black-box uncertainty signal that instantiates selective classification and distribution-free calibration for agentic systems. Across 8,000 runs of four models on 200 HotpotQA questions, consistent tasks (at most 2 unique paths) achieve 82-87% accuracy while inconsistent tasks (4 or more paths) achieve 41-65%, a gap that survives controls for task difficulty. Divergence concentrates at step 2 (50.5% of Llama tasks), and consistency metrics detect failures with AUROC 0.62-0.78. Exploiting this signal, selective prediction (answering only when k=3 runs agree) achieves 87-88% accuracy at 54-62% coverage, a 6-14pp gain over single-run baselines, and matches a split-conformal baseline without a held-out calibration set. A cross-benchmark validation on SWE-bench (50 tasks, 1,000 runs) preserves the consistency hierarchy while revealing an ~8x spread in mean trajectory length across models, and bootstrap analysis shows single-run evaluations misrank models 29.3% of the time.

Safety alignment in Large Language Models is critical for healthcare; however, reliance on binary refusal boundaries often results in over-refusal of benign queries or unsafe compliance with harmful ones. While existing benchmarks measure these extremes, they fail to evaluate Safe Completion: the model's ability to maximise helpfulness on dual-use or borderline queries by providing safe, high-level guidance without crossing into actionable harm. We introduce Health-ORSC-Bench, the first large-scale benchmark designed to systematically measure Over-Refusal and Safe Completion quality in healthcare. Comprising 31,920 benign boundary prompts across seven health categories (e.g., self-harm, medical misinformation), our framework uses an automated pipeline with human validation to test models at varying levels of intent ambiguity. We evaluate 30 state-of-the-art LLMs, including GPT-5 and Claude-4, revealing a significant tension: safety-optimised models frequently refuse up to 80% of "Hard" benign prompts, while domain-specific models often sacrifice safety for utility. Our findings demonstrate that model family and size significantly influence calibration: larger frontier models (e.g., GPT-5, Llama-4) exhibit "safety-pessimism" and higher over-refusal than smaller or MoE-based counterparts (e.g., Qwen-3-Next), highlighting that current LLMs struggle to balance refusal and compliance. Health-ORSC-Bench provides a rigorous standard for calibrating the next generation of medical AI assistants toward nuanced, safe, and helpful completions. Furthermore, our benchmark facilitates reproducible evaluation, encourages safety calibration, and supports development of clinically reliable, context-aware, human-aligned medical AI systems. Our code and data are available at: https://github.com/ZhihaoZhang97/Health-ORSC-Bench. Warning: Some contents may include toxic or undesired contents.

Building NLP systems for subjective tasks requires one to ensure their alignment to contrasting human values. We propose the MultiCalibrated Subjective Task Learner framework (MC-STL), which clusters annotations into identifiable human value clusters by three approaches (similarity of annotator rationales, expert-value taxonomies or rater's sociocultural descriptors) and calibrates predictions for each value cluster by learning cluster-specific embeddings. We demonstrate MC-STL on several subjective learning settings, including ordinal, binary, and preference learning predictions, and evaluate it on multiple datasets covering toxic chatbot conversations, offensive social media posts, and human preference alignment. The results show that MC-STL consistently outperforms the baselines that ignore the latent value structure of the annotations, delivering gains in discrimination, value-specific calibration, and disagreement-aware metrics.

Background: Smartphone-based dermatology requires inter-device colorimetric reliability that holds across calibration regimes, yet quantitative multi-device benchmarks remain scarce. Materials and Methods: We analyzed matched facial images from 965 Korean subjects captured by a digital single-lens reflex (DSLR) camera, a consumer tablet, and a consumer smartphone, and evaluated two calibration methods against the DSLR reference. The methods are standard global linear Color Correction Matrix (CCM) normalization and region-specific CCM trained per anatomical region, both applied in Commission Internationale de l'Eclairage Lab* (CIELAB) space. Results: Linear CCM reduced inter-device color differences by 61-74% and placed both Melanin Index (intraclass correlation coefficient [ICC] = 0.80) and Individual Typology Angle (ITA, ICC = 0.78) in the good reliability band. Region-specific CCM raised both indices into the excellent reliability band (MI ICC = 0.95, ITA ICC = 0.93), with anatomical region exceeding the source device as the largest pre-calibration variance contributor (analysis-of-variance $η^2 = 0.18$ versus 0.12). Conclusion: Consumer-device skin colorimetry therefore achieves clinically useful inter-device reliability using standard calibration, with region-aware calibration the largest remaining source of improvement.

Vision-language models (VLMs) such as CLIP are increasingly adapted across decentralized data silos, yet the reliability of their predictions under federated learning (FL) remains largely unexplored. In this work, we present a systematic study of calibration in federated CLIP under non-IID client distributions. Our experiments reveal that widely used prompt-tuning methods consistently degrade calibration, often yielding substantially higher calibration error despite competitive recognition performance, while explicit training-time calibration regularizers provide only limited improvements. Motivated by these findings, we identify the choice of fine-tuning parameterization as a critical factor governing calibration and conduct a controlled comparison between prompt tuning and five backbone fine-tuning (BFT) strategies: AdaptFormer, LayerNorm, LoRA, VeRA, and DoRA. We find that BFT methods generally offer a more favorable accuracy-calibration trade-off than prompt tuning, although their benefits are not universal. Through extensive analysis, we show that calibration behavior is closely linked to the geometry of federated updates, residual parameterization, and the resulting client and logit drift. Across in-distribution, domain-generalization, and base-to-new evaluation settings, our results establish fine-tuning parameterization as a central design choice for building accurate and reliable federated CLIP models. Codes are available at https://github.com/mainaksingha01/FL2oRA.

Accurate intrinsic calibration is a crucial yet challenging prerequisite for 3D reconstruction using light field cameras. Existing calibration models typically analyze the main lens and micro lens array (MLA) in a coupled manner, resulting in high complexity and a large number of parameters. In this paper, we propose a linear fractional transformation (LFT) model that introduces a single parameter $α$ to decouple the imaging processes of the main lens and the MLA. A dedicated matrix $\mathbf{H}_α$ is designed to characterize the MLA projection, enabling the main lens and the MLA to be calibrated independently. The proposed calibration method consists of an analytical least-squares solution for $\mathbf{H}_α$, followed by joint nonlinear refinement of all intrinsic parameters. Experimental results on both physical datasets and simulated data demonstrate that the proposed method achieves a mean translation error of $2.1\%$, outperforming the state-of-the-art, while maintaining sub-pixel reprojection accuracy. The complete codebase, including a light field simulator based on the proposed model, is openly available to the research community.

Tactile sensing plays a key role in enabling dexterous and reliable robotic manipulation, but realizing this capability requires substantial calibration to convert raw sensor readings into physically meaningful quantities. Despite its near-universal necessity, the calibration process remains ad hoc and labor-intensive. Here, we introduce 3D Cal, an open-source library that transforms a low-cost 3D printer into an automated probing device capable of generating large volumes of labeled training data for calibrating vision-based tactile sensors. 3D Cal also provides an end-to-end, user-friendly pipeline for training custom convolutional networks to produce high-quality depth reconstructions. Using 3D Cal, we systematically explore the relationship between training data volume and spatial reconstruction performance on two commercially available sensors, DIGIT and GelSight Mini, and derive practical, empirically-grounded guidelines for calibrating these sensors. Finally, we demonstrate depth reconstruction performance on the DIGIT and GelSight Mini comparable to state-of-the-art methods, achieving average reconstruction errors of 156 $\mathrm{μm}$ and 205 $\mathrm{μm}$ on unseen objects, respectively. By automating tactile sensor calibration, 3D Cal can accelerate tactile sensing research, simplify sensor deployment, and facilitate the integration of tactile sensing in robotic platforms.

The "alignment tax" of post-training is typically framed as a drop in task accuracy. We show it also involves a severe loss of calibration, making models overconfident, less reliable, and model outputs less diverse. We show that this trade-off can be navigated effectively via a simple post-hoc intervention: interpolating between a model's weights before and after alignment. Crucially, this is not a strict trade-off. We find that the process consistently reveals Pareto-optimal interpolations - models that improve accuracy beyond both parents while substantially recovering the calibration lost during alignment. Our work demonstrates that simple model merging provides a computationally efficient method for mitigating the full scope of the alignment tax, yielding models that are more capable and more reliable.

Confidence calibration, the alignment of a model's predicted confidence with its actual accuracy, is crucial for the reliable deployment of Large Language Models (LLMs). However, this critical property remains largely under-explored in multilingual contexts. In this work, we conduct the first large-scale, systematic studies of multilingual calibration across six model families and over 100 languages, revealing that non-English languages suffer from systematically worse calibration. To diagnose this, we investigate the model's internal representations and find that the final layer, biased by English-centric training, provides a poor signal for multilingual confidence. In contrast, our layer-wise analysis uncovers a key insight that late-intermediate layers consistently offer a more reliable and better-calibrated signal. Building on this, we introduce a suite of training-free methods, including Language-Aware Confidence Ensemble (LACE), which adaptively selects an optimal ensemble of layers for each specific language. Our study highlights the hidden costs of English-centric alignment and offer a new path toward building more globally equitable and trustworthy LLMs by looking beyond the final layer.

Active learning (AL) aims to build high-quality labeled datasets by iteratively selecting the most informative samples from an unlabeled pool under limited annotation budgets. However, in fine-grained image classification, assessing this informativeness reliably is especially challenging due to subtle differences between classes. In this paper, we introduce a novel active learning method, combining discrepancy-confusion uncertainty and calibration diversity for active fine-grained image classification (DECERN), to effectively perceive the distinctiveness between fine-grained images and evaluate the sample value. DECERN introduces a multifaceted informativeness measure that combines discrepancy-confusion uncertainty and calibration diversity. The discrepancy-confusion uncertainty quantifies the structural stability and category directionality of fine-grained unlabeled data during local feature fusion. Subsequently, uncertainty-weighted clustering is performed to diversify the uncertainty samples. Then we calibrate the diversity to maximize the global diversity of the selected sample while maintaining its local representativeness. Extensive experiments conducted on 7 fine-grained image datasets across 39 distinct experimental settings demonstrate that our method achieves superior performance compared to state-of-the-art methods.

Fairness audits are a key component of responsible machine-learning deployment. Yet, audit-recommendation reliability under incomplete protected-label access is still poorly understood. In this work, we focused on protected-label missingness in fairness mitigation audits. We introduced a seed-calibrated stress test to separate missingness effects from seed-to-seed movement already present under complete labels. Across ACS/Folktables tasks, missingness settings that retain some protected labels usually do not move selected mitigation methods beyond a complete-label seed-to-seed baseline. At $0$ protected-label access, candidates collapse to an empirical-risk-minimization baseline and deterministic tie-breaking rather than revealing a broad missingness effect. We also found that threshold optimization can turn fairness gains on a single protected axis into intersectional harm above a seed baseline, and this threshold-optimizer finding persists under random-forest validation. Overall, our results highlight that protected-label missingness should be reported with seed-null calibration, candidate-set context, and intersectional consequences before it is treated as evidence of audit fragility.

Deep neural networks have been increasingly used in safety-critical applications such as medical diagnosis and autonomous driving. However, many studies suggest that they are prone to being poorly calibrated and have a propensity for overconfidence, which may have disastrous consequences. In this paper, unlike standard training such as stochastic gradient descent, we show that the recently proposed sharpness-aware minimization (SAM) counteracts this tendency towards overconfidence. The theoretical analysis suggests that SAM allows us to learn models that are already well-calibrated by implicitly maximizing the entropy of the predictive distribution. Inspired by this finding, we further propose a variant of SAM, coined as CSAM, to ameliorate model calibration. Extensive experiments on various datasets, including ImageNet-1K, demonstrate the benefits of SAM in reducing calibration error. Meanwhile, CSAM performs even better than SAM and consistently achieves lower calibration error than other approaches

Reliable probability estimates by classifiers are essential in high-risk applications. In practice, however, predicted probabilities are often miscalibrated, and many existing post-hoc calibration methods typically lack guarantees that a specific notion of calibration is achieved after the correction procedure is applied. We introduce a set-based perspective on calibration through the notion of cumulative mass calibration and the corresponding error measures. We propose a new calibration procedure based on conformal prediction that forms cumulative probabilities with guaranteed marginal coverage. We introduce an adaptive temperature scaling algorithm, with the temperature tuned for each input to satisfy the conformal coverage constraint. As we show, this procedure can be efficiently implemented. Across image classification tasks, particularly in settings with many classes, our method improves newly introduced calibration error measures (CMCE and $α$-CMCE) and standard metrics (such as ECE, cw-ECE, MCE) over the existing baselines.

This work presents an RGB-D imaging-based approach to marker-free hand-eye calibration using a novel implementation of the iterative closest point (ICP) algorithm with a robust point-to-plane (PTP) objective formulated on a Lie algebra. Its applicability is demonstrated through comprehensive experiments using three well known serial manipulators and two RGB-D cameras. With only three randomly chosen robot configurations, our approach achieves approximately 90% successful calibrations, demonstrating 2-3x higher convergence rates to the global optimum compared to both marker-based and marker-free baselines. We also report 2 orders of magnitude faster convergence time (0.8 +/- 0.4 s) for 9 robot configurations over other marker-free methods. Our method exhibits significantly improved accuracy (5 mm in task space) over classical approaches (7 mm in task space) whilst being marker-free. The benchmarking dataset and code are open sourced under Apache 2.0 License, and a ROS 2 integration with robot abstraction is provided to facilitate deployment.

We study the fundamental problem of calibrating a linear binary classifier of the form $σ(\hat{w}^\top x)$, where the feature vector $x$ is Gaussian, $σ$ is a link function, and $\hat{w}$ is an estimator of the true linear weight $w^\star$. By interpolating with a noninformative $\textit{chance classifier}$, we construct a well-calibrated predictor whose interpolation weight depends on the angle $\angle(\hat{w}, w_\star)$ between the estimator $\hat{w}$ and the true linear weight $w_\star$. We establish that this angular calibration approach is provably well-calibrated in a high-dimensional regime where the number of samples and features both diverge, at a comparable rate. The angle $\angle(\hat{w}, w_\star)$ can be consistently estimated. Furthermore, the resulting predictor is uniquely $\textit{Bregman-optimal}$, minimizing the Bregman divergence to the true label distribution within a suitable class of calibrated predictors. Our work is the first to provide a calibration strategy that satisfies both calibration and optimality properties provably in high dimensions. Additionally, we identify conditions under which a classical Platt-scaling predictor converges to our Bregman-optimal calibrated solution. Thus, Platt-scaling also inherits these desirable properties provably in high dimensions.

Deploying deep learning-based imaging tools across various clinical sites poses significant challenges due to inherent domain shifts and regulatory hurdles associated with site-specific fine-tuning. For histopathology, stain normalization techniques can mitigate discrepancies, but they often fall short of eliminating inter-site variations. Therefore, we present Data Alchemy, an explainable stain normalization method combined with test time data calibration via a template learning framework to overcome barriers in cross-site analysis. Data Alchemy handles shifts inherent to multi-site data and minimizes them without needing to change the weights of the normalization or classifier networks. Our approach extends to unseen sites in various clinical settings where data domain discrepancies are unknown. Extensive experiments highlight the efficacy of our framework in tumor classification in hematoxylin and eosin-stained patches. Our explainable normalization method boosts classification tasks' area under the precision-recall curve(AUPR) by 0.165, 0.545 to 0.710. Additionally, Data Alchemy further reduces the multisite classification domain gap, by improving the 0.710 AUPR an additional 0.142, elevating classification performance further to 0.852, from 0.545. Our Data Alchemy framework can popularize precision medicine with minimal operational overhead by allowing for the seamless integration of pre-trained deep learning-based clinical tools across multiple sites.

A set of probabilistic forecasts is calibrated if each prediction of the forecaster closely approximates the empirical distribution of outcomes on the subset of timesteps where that prediction was made. We study the fundamental problem of online calibrated forecasting of binary sequences under the standard $\ell_1$ calibration error metric, which was initially studied by Foster & Vohra (1998). They derived an algorithm with $O(T^{2/3})$ calibration error after $T$ time steps, and showed a lower bound of $Ω(T^{1/2})$. These bounds remained stagnant for two decades, until Qiao & Valiant (2021) improved the lower bound to $Ω(T^{0.528})$ by introducing a combinatorial game called sign preservation and showing that lower bounds for this game imply lower bounds for calibration. In this paper, we give the first improvement to the $O(T^{2/3})$ upper bound on calibration error of Foster & Vohra. We do this by introducing a variant of Qiao & Valiant's game that we call sign preservation with reuse (SPR). We prove that the relationship between SPR and calibrated forecasting is bidirectional: not only do lower bounds for SPR translate into lower bounds for calibration, but algorithms for SPR also translate into new algorithms for calibrated forecasting. We then give an improved upper bound for the SPR game, which implies, via our equivalence, a forecasting algorithm with calibration error $O(T^{2/3 - \varepsilon})$ for some $\varepsilon > 0$, improving Foster & Vohra's upper bound for the first time. Using similar ideas, we then prove a slightly stronger lower bound than that of Qiao & Valiant, namely $Ω(T^{0.54389})$. Our lower bound is obtained by an oblivious adversary, marking the first $ω(T^{1/2})$ calibration lower bound for oblivious adversaries.

We introduce \emph{Calibrated Stackelberg Games (CSGs)}, a generalization of the standard Stackelberg Games (SGs) framework. In CSGs, a principal repeatedly interacts with an agent who (contrary to standard SGs) does not have direct access to the principal's action but instead best-responds to calibrated forecasts about it. This framework provides a powerful and realistic modeling tool that goes beyond assuming that agents use ad hoc and highly specified algorithms for interacting in strategic settings and instead builds on statistical foundations of forecasts and calibration. We show that in CSGs, despite both the principal and the agent having less information than in standard SGs, the principal's optimal utility remains upper and lower bounded by the Stackelberg value of the one-shot game, in both finite and continuous settings. Alongside CSGs, we develop stronger notions of calibration and corresponding algorithms that address two central challenges for calibration in game-theoretic environments. First, achieving point-wise calibration typically incurs an error that scales exponentially with the dimension of the strategy space. Second, the principal's convergence rate in CSGs depends critically on the adaptivity of the agent's calibration algorithm. To address these challenges, we establish a meaningful, efficiently achievable relaxation of calibration based on conditioning on best-response regions. This yields the first notion of calibration in games with a statistical rate that only depends on the number of agents' actions rather than the dimension of the principal's strategy space and that leads to no-swap regret for the agent. We further develop adaptive calibration algorithms for the agents that provide fine-grained, any-time calibration guarantees against adversarial sequences, enabling the principal to achieve faster convergence in CSGs.

Feed-forward 3D reconstruction models output a per-pixel confidence that is used by downstream systems as an uncertainty signal. The confidence is trained to serve as a weight in the training loss of models. Whether the confidence can be used as an uncertainty magnitude has not been measured. We audit seven backbones on 13 datasets and score the confidence on four properties, i.e., ranking of error, ratio of error to uncertainty on average, slope of this ratio across the confidence range, and coverage of the implied error distribution. Although the confidence ranks error quite well, the uncertainty decoded from the confidence is too small compared to the actual error. The uncertainty has the right size only under the exact training conditions. The median case is off by at least 2.4x across all seven models, while the uncertainty is further off the more confident the model is. Our work shows that the overconfidence appears on unseen scenes even when the model reaches its loss's optimum. As a post-hoc repair we fit a power law on the confidence with two constants per backbone--dataset pair. The repair brings all four audited properties to target at the dataset level, while leaving ranking untouched. Fitted with the target dataset held out, the constants bring the median case from 2.4x off to 1.35x. The repair does not hold below the dataset level, where two-thirds of held-out scenes are still more than five points off in coverage. We attribute what the repair cannot reach to the model, which carries neither the scale of the error nor the shape of its distribution across predictions. We release the audit protocol, its results, and the fitted constants per backbone-dataset pair.

Recent work has argued normatively, on synthetic data, that evaluating survival models by discrimination alone (concordance index) yields systematically misleading model comparisons, because the metric ignores calibration and time-dependent accuracy. Whether this matters for real, published, non-clinical models has not been tested. We reproduce three published survival-ML models across three structurally distinct domains -- hard-drive failure prediction, peer-to-peer credit default, and user disengagement on digital platforms -- validate our instrument against the anchor paper's own synthetic experiment, and test five pre-registered hypotheses under a Holm-corrected family-wise error rate. Three of five reject (though one pre-registered threshold clears by a narrow margin). A model reproducing the published literature's discrimination almost exactly (C = 0.9595 vs. 0.958 reported) fails a formal calibration test at p < 0.001; a broad feature-ablation search finds no single attribute responsible for its discrimination, so the calibration failure is not a trivial shortcut artifact. A lender's estimated default risk is biased upward by roughly two percentage points, growing to nearly four in the riskiest segment, when loan prepayment is treated as non-informative censoring rather than a competing risk. A platform's churn model shows probability estimates that degrade with the horizon even as global discrimination stays within the pre-registered C-index band. A direct test of whether metric choice inverts model preference does not reject, though with limited power given two to three models per domain; the failure mode we document is better characterized as misplaced confidence in a chosen model than as choosing the wrong one. We release a pre-registered evaluation harness with full code and an annotated notebook, so these results can be verified independently and the audit extended.