VLM Calibration

VLM: Vision-Language Model

Follow this topic Ask about this topic

Momentum

13 papers in the last four weeks, up 30% on the four weeks before. 0.1% of all new papers.

Jul 13Week of Sep 28

Latest papers 47

Oct 1, 2026cs.AI

Towards Reliable Vision-Language Models for Autonomous Driving

Vision-Language models (VLMs) are increasingly being explored in autonomous driving for tasks such as scene understanding, driving reasoning, decision-making, and end-to-end driving. As their role becomes more prominent, ensuring their robustness and reliability is increasingly important. In real-world conditions, visual inputs may be degraded by sensor imperfections and environmental conditions, potentially affecting both model predictions and their associated confidence. Such degradation is especially concerning in autonomous driving, where safety-critical decisions require models to make accurate predictions and recognize when their predictions may be unreliable. In this work, we evaluate five VLMs (Qwen3.5-9B, Gemma4-E4B, LLaVA-OneVision-7B, DriveFusion/DriveFusionQA-4B, and NVIDIA Alpamayo-1.5-10B) across four driving-related QA datasets with different visual input settings, including single-frame, multi-view, multi-frame, and monocular inputs. Our results show that the effects of visual corruption vary across models, datasets, and input settings, with changes in accuracy and confidence reliability and also differing across conditions. We then apply Visual Evidence Augmentation (

\mathrm{V}{\scriptstyle \mathrm{EA}}

), a recent inference-time method to examine whether it can improve model reliability under degraded visual conditions. We find that

\mathrm{V}{\scriptstyle \mathrm{EA}}

improves performance for some models and datasets, although the gains are not consistent across all settings.

Manasa Mariam Mammen, Priyanka Mary Mammen, Zafer Kayatas +1

Technical University of Munich, Heilbronn, Germany · Mercedes Benz AG, Sindelfingen, Germany · University of Massachusetts, Amherst, USA

VLM Robustness Image Corruption Robustness VLMs for Autonomous Driving VLM Calibration

Sep 28, 2026cs.CV

SubRot: Signed Gradient Subspace Calibration for VLM Rotation Quantization

Post-training quantization reduces the deployment cost of vision-language models (VLMs), but preserving multimodal capabilities at low bit widths remains challenging. Existing methods rely on modality- or token-level gradient statistics, which are susceptible to cross-sample variations in visual-to-textual token ratios and the positions of visual information, limiting statistical stability. Moreover, overly coarse aggregation through absolute values and averaging discards gradient signs and channel-wise differences, limiting the separation of modality-specific sensitivities. In contrast, the channel space provides a shared coordinate system across samples, making it a more natural basis for capturing stable task-sensitive structures. We therefore propose SubRot, a signed gradient subspace calibration method for VLM rotation quantization. Through eigendecomposition of the empirical Fisher matrix of activation gradients, SubRot identifies a sensitive channel subspace with three properties: cross-sample stability, clear sensitivity separation, and consistent signed effects on the autoregressive loss along certain directions. Guided by a local Taylor expansion, SubRot combines signed first-order guidance along sign-stable directions with second-order constraints along the remaining sensitive directions, while retaining MSE for overall reconstruction quality. This objective steers quantization errors toward loss-decreasing directions while controlling their magnitude. Experiments on five VLMs across five benchmarks show consistent average-score improvements over FlatQuant under W4A6 and W4A4, reaching 1.4 percentage points on LLaVA-NeXT-7B. Under W4A4, average accuracy degradation from FP16 remains within 1.4 percentage points across all evaluated models, while LLaVA-v1.5-13B exceeds its FP16 average score by 0.4 percentage points.

Zhenhao Shang, Haizhao Jing, Haokui Zhang +4

Northwest Polytechnical University · Nanjing University of Science and Technology · Intellifusion +1

VLM Quantization Rotation-Based Quantization VLM Calibration Post-Training Quantization

Sep 24, 2026cs.CV

Domain Recentering and Confidence-Weighted Prior Calibration for Vision-Language Models

Vision-language models such as CLIP achieve strong zero-shot classification, yet under distribution shift, visual embeddings drift from fixed text embeddings. Training-free calibration avoids the per-sample optimization of prompt learning, but prior feature calibration gives each image the full bias of one hard cluster. We propose Domain Recentering with Confidence Calibration (DRC), a training-free method adapting CLIP from a set of unlabeled target images. DRC fits a Gaussian mixture once and subtracts from each embedding a posterior-weighted average of component means. It then removes residual class preference with a log-prior correction, estimating the prior from confidence-weighted predictions. Among compared methods, DRC achieves the highest average accuracy on cross-domain datasets, exceeding zero-shot CLIP by 4.13 and 5.07 points with ViT-B/16 and ResNet-50, with gains over CLIP also holding under ImageNet distribution shifts.

Youngeun Seol, Jimin Shin, Heeseo Yoon +1

Department of Computer Science and Engineering, Ewha Womans University

Source-Free Domain Adaptation Unsupervised Domain Adaptation Test-Time Adaptation of VLMs VLM Calibration

Sep 17, 2026cs.AI

Perception, Layout, and Validation: Calibrated Confidence for Reliable Straight-Through Processing of Financial Documents

Straight-through processing (STP) on extracted key-value fields from financial documents without human review requires a calibrated probability together with a bounded guarantee on the residual error of the auto-approved tier. The emergence of modern Vision Language Models (VLMs) provides an out-of-the-box capability for extracting the key-values, but their verbalized confidence signals are unreliable and weakly track field correctness. This paper introduces a decomposed confidence layer along three interpretable channels, including perception, layout, and validation. Together with a final conformal risk control, the score can be used for reliable STP of financial documents. The method is validated on three public datasets covering real invoices, synthetic invoices, and ad-buy forms, using two different VLM families (Qwen3.6-27B and Gemini-3.1-Flash-Lite). Our decomposed score consistently improves the separation of correct from incorrect extractions, substantially raising the AUROC from 0.54-0.74 for VLM verbalized signals to 0.90-0.99 with contributions from all three designed channels. Crucially for industrial deployment, this enables usable STP. The native VLM confidence signals could clear only 0.1%-7.0% of fields under risk control at a target error of <10%. In contrast, the proposed method auto-approves 49-72% of fields while holding the empirical error of the accepted tier at or below the target.

Yichao Jin, Yushuo Wang, Yuxuan Han +7

OCBC, Singapore

Document Information Extraction VLM Calibration Document Image Understanding

Sep 16, 2026cs.RO

Calibrated Probabilistic Obstruction Reasoning with Vision-Language Models for Grasping in Clutter

Retrieving a target from clutter requires deciding whether to grasp the target, remove a blocker, or defer. Existing methods typically commit to a single obstruction graph or removal strategy, ignoring uncertainty across alternative scene interpretations. They also rely on miscalibrated vision-language model (VLM) predictions and can produce pairwise obstruction relations that are jointly inconsistent. Moreover, current approximations provide no guarantees about the impact of discarded hypotheses on the final decision. We propose CPOR-Grasp, a calibrated probabilistic obstruction-reasoning framework that propagates uncertainty from pairwise evidence to action decisions. CPOR-Grasp calibrates and fuses VLM, depth, and amodal-mask cues to estimate obstruction probabilities, induces a distribution over valid obstruction graphs, and marginalizes over these graphs to compute the likelihood that the target is accessible or that a given blocker should be removed. To make inference tractable, it retains only the highest-probability graphs and derives a total-variation bound on the discarded probability mass, enabling certified decisions, adaptive stopping, and principled deferral. On synthetic and real UNOBench scenes, CPOR-Grasp outperforms state-of-the-art baselines. Calibration error decreases from 0.1416 to 0.0185 on the Gemini Robotics backbone, while graph truncation matches exact inference on 99.74% of decisions using 56 times fewer graphs. In real-world experiments, CPOR-Grasp achieves a 77.8% average success rate, surpassing SOTA baselines.

Thanh-Tuan Tran, Ngoc-Chien Chu, Thanh Nguyen Canh +3

University of Engineering and Technology, Vietnam National University, Hanoi, 10000, Vietnam · School of Information Science, Japan Advanced Institute of Science and Technology, Ishikawa, 923-1211, Japan · Department of Robotics, Hanyang University, Gyeonggi-do, 15588, Korea

Robotic Grasping Robotic Manipulation VLM Calibration VLMs for Robotics

Sep 16, 2026cs.AI

The Mirage of Calibrated Confidence: Trajectory-Independence of Verbalized Confidence in Vision-Language Models

A calibrated Vision-Language Model (VLM) can repeatedly self-correct, say "Wait, I should recheck," arrive at the wrong answer, and still report high confidence. We find that this occurs because verbalized confidence is largely trajectory-independent in the VLMs and calibration methods we evaluate. We examine this through three complementary lenses: content variation, token masking, and the model's own hesitation markers. We show that confidence is insufficiently sensitive to what the reasoning trajectory actually contains, and that calibration training can paradoxically worsen this disconnect. Since existing metrics like ECE and AUROC cannot detect this problem, we propose the Trajectory-Grounding Score (TGS) in two complementary forms: TGS-self, which compares confidence with and without access to the model's own trajectory, and TGS-pair, which tests whether the model assigns higher confidence to correct trajectories than to flawed ones along the vision, reasoning, and answer axes. We propose TGS-Bench, a model-agnostic suite spanning 10 benchmarks with controlled good/bad trajectory pairs, and show that conventional calibration rankings diverge from trajectory-grounding rankings, exposing a blind spot in current evaluation practice.

Jisoo Yang, Jaeho Han, Trung X. Pham +1

Department of Artificial Intelligence, Chung-Ang University, Republic of Korea · Van Lang University, Ho Chi Minh City, Vietnam

VLM Evaluation Confidence Estimation in Language Models VLM Reasoning VLM Calibration

Sep 15, 2026cs.CV

Can VLMs Reliably Assess Sidewalk Accessibility Attributes from Pedestrian-Level Imagery?

An important component of urban accessibility, particularly for wheelchair users and people with reduced mobility, is sidewalk compliance with measurable requirements. We test whether effective width, longitudinal slope, cross slope, and pavement condition can be assessed reliably from pedestrian-level imagery using vision-language models (VLMs). We present the first application of sampling-based conformal prediction (CP) for VLM-based accessibility assessment. We evaluate four VLMs on 514 sidewalk images from Seoul, South Korea, with field-measured ground truth. Conformal calibration attains the nominal 90% coverage for all models and attributes, but the calibrated regions differ in informativeness. Effective width yields the most informative estimates, with a mean interval half-width of about 1.0 m for the best model. Since every model overestimates width, asymmetric calibration shortens the intervals by up to 33% at unchanged coverage. Longitudinal slope is marginally informative, cross-slope intervals are too wide to resolve regulatory thresholds, and pavement-condition sets degenerate to all five grades (A-E) for three of the four models. Uncalibrated intervals from raw sampling dispersion cover only 17-47% of field-measured values at a nominal 90% level. Among the images with the most self-consistent responses, these intervals miss the field-measured value in up to 96% of cases. Response self-consistency is therefore not evidence of accuracy, and sampling dispersion cannot be interpreted as uncertainty until it has been calibrated against field-measured ground truth. No quantitative attribute reaches the precision required for general compliance assessment, but CP identifies from calibration data alone which attributes can support screening of segments far from the thresholds. We release the annotated pedestrian-level images and their corresponding field-measured attribute values.

Seung Jae Lieu, Diego Morra, Chiara Cadoni +3

Senseable City Lab Massachusetts Institute of Technology Cambridge, MA 02139

VLM Evaluation Urban Planning VLM Calibration

Sep 15, 2026cs.LG

Bridging the Confidence Gap: Temperature Scaling for Calibrating Test-Time Prompt Tuning

Test-time prompt tuning (TPT) enables adaptation on a single test instance, achieving improved accuracy but often sacrificing calibration performance. Most existing calibration methods introduce additional regularization terms to promote dispersion across text embeddings and reduce calibration error, yet these methods often suffer from a drop in accuracy. Motivated by the well-calibrated nature of zero-shot predictions, we propose CoTS, a simple yet effective post-hoc calibration method that preserves accuracy. Specifically, CoTS applies temperature scaling to minimize the confidence gap between adapted and zero-shot predictions. To fully exploit the potential of multiple augmentations during adaptation, we introduce a weak-strong ensemble strategy that further boosts accuracy. We then apply CoTS to this ensemble, termed E-CoTS, to maintain its well-calibrated property. Extensive experiments on diverse datasets and backbones show that our approaches effectively mitigate miscalibration without compromising primary accuracy. For instance, E-CoTS reduces the average expected calibration error of TPT from 11.90% to 5.38% on ImageNet variants, while even increasing accuracy from 60.74% to 62.95%. Moreover, when integrated with existing calibration methods, E-CoTS usually enhances both accuracy and calibration simultaneously.

Yuwei Liang, Jian Liang, Dapeng Hu +2

School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing, China · NLPR & MAIS, Institute of Automation, Chinese Academy of Sciences, Beijing, China · Independent Researcher

Post-Hoc Calibration Test-Time Adaptation of VLMs Model Calibration VLM Calibration Test-Time Prompt Tuning

Sep 15, 2026cs.CV

SAVOR: Self-Aware Visual Grounding via Confidence-Calibrated Reinforcement Learning for Multimodal Hallucination Mitigation

Multimodal large language models (MLLMs) have made strong progress on visual question answering and image captioning, yet they still produce fluent claims about objects, attributes, or relations that are not grounded in the image. Many remedies either modify decoding at test time, which adds latency, or fine tune with preferences such as DPO variants, which teach which answer is preferred but not when the model's own answer is unreliable. We argue that calibrated self assessment is the missing signal. We introduce Savor, a training framework that (i) augments the output schema with token and answer confidence, (ii) optimises the policy with a Group Relative Policy Optimisation (GRPO) objective that penalises calibration error and poor abstention decisions, and (iii) uses the learned confidence at inference time to revisit visual evidence only when the model is uncertain. Experiments on POPE, HallusionBench, AMBER and MMHal-Bench across two recent backbones (InternVL3-8B and Qwen3-VL-8B) show that Savor reduces hallucination while preserving general capability on MME and MMBench, with lower Expected Calibration Error than DPO and decoding baselines.

Zixiu Ding, Zilin Zhao, Yingjie He +3

Central South University · Peking University

Object Hallucination in VLMs Visual Grounding VLM Calibration VLM Hallucination Mitigation RL for VLMs

Sep 14, 2026cs.CV

Human-Grounded Calibration for Long-Text Image-Text Congruence in Vision-Language Models

Long-text image--text congruence scoring is increasingly important for vision-language systems that must evaluate whether detailed textual descriptions match visual content. However, raw similarity scores from dual-encoder models are difficult to interpret as calibrated congruence measures, especially under the modality gap between image and text embeddings. This paper proposes Congruency Score (CS), a lightweight calibration layer that maps image--text similarity evidence into a bounded score. Using DOCCI and Urban1k, we evaluate four frozen vision-language backbones and show that observed reductions in post-projection centroid distance do not uniformly improve image--text retrieval performance. Human-grounded evaluations on DOCCI further reveal a trade-off: direct post-hoc calibration preserves high association with human judgments, whereas selected projection-based configurations can reduce threshold-relevant slope and intercept distortions at the cost of retrieval performance and association strength. These results establish long-text image--text congruence scoring as a calibrated score-estimation problem, where retrieval performance, human association, and threshold calibration must be evaluated as distinct objectives. CS provides a lightweight way to expose and operationalize this separation.

Alessandro Gambetti, Qiwei Han

NOVA School of Science and Technology · Universidade NOVA de Lisboa · Caparica, Portugal +2

Text-Image Alignment Vision-Language Retrieval VLM Calibration

Sep 13, 2026cs.CV

Vision-Language Models for Criterion-Level Grading of Handwritten Examinations in Outcome-Based Education

Criterion-level grading connects examination performance to learning outcomes, but manual marking introduces workload and variation between markers. This study evaluates vision-language models (VLMs) for handwritten outcome-based assessment across five dimensions: accuracy, human agreement, repeated-run reliability, error concentration, and explanation quality. Using 1,982 criterion-level records from 485 undergraduate examination answers, we compare 20 configurations spanning Qwen2.5-VL, InternVL3, Pixtral, a Donut baseline, and a cascade ensemble. Evaluation setups include zero-shot prompting, few-shot prompting, partial fine-tuning, and Low-Rank Adaptation (LoRA). Two independent faculty markers regraded all 291 test criteria, providing a human agreement baseline on the same assessment materials. Qwen2.5-VL with LoRA achieved Quadratic Weighted Kappa (QWK) of 0.727 and mean absolute error of 0.435 marks against the examiner, compared with mean human-pair QWK of 0.551. This comparison reflects calibration to the examiner's training marks. LoRA outperformed partial fine-tuning for all three instruction-tuned VLMs, while few-shot prompting reduced QWK in every configuration with valid prompted scores. Aggregate reliability and exact repeatability diverged: intraclass correlations ranged from 0.790 to 0.874, yet 50.2-63.6% of criteria changed marks across five sampled runs. Attention-guided deletion showed no statistically significant advantage over random masking, and four faculty reviewers reached no consensus on explanation usefulness. These findings highlight the need for rubric-specific calibration, repeatable scoring, review of consequential errors, and separate validation of explanations. The released evaluation protocol supports criterion-level assessment research and grading tools with teacher oversight.

Asif Hasan Tonmoy, Saad Ahmed, Md Khalid Syfullah +1

VLM Evaluation Rubric-Based Evaluation VLM Calibration Automated Grading

Sep 9, 2026cs.CV

Learning to Adapt and Calibrate: Score Distribution Alignment for Few-Shot Uncertainty Prediction in Medical VLMs

Uncertainty estimation for medical vision--language models (VLMs) using conformal prediction has gained increasing attention due to its distribution-free coverage guarantees. However, standard conformal prediction relies on exchangeability between calibration and test data and typically requires a sufficiently large calibration set to obtain reliable coverage. These assumptions are difficult to satisfy in few-shot transfer settings, where only a small labeled support set is available to adapt a pretrained VLM to a new medical task, while an unlabeled query set is used for evaluation. Supervised fine-tuning on the support set changes the model parameters and consequently shifts the nonconformity score distribution, breaking exchangeability between calibration and query samples and leading to unreliable coverage under distribution shift. Existing transductive conformal adaptation methods often preserve validity by avoiding supervised updates. While this helps maintain conformal assumptions, it underutilizes the scarce labeled support data and limits task adaptation, which is the primary objective in few-shot learning. In this setting, conformal prediction should serve as an uncertainty estimation layer that supports the adapted model, rather than preventing adaptation itself. To this end, we propose AlignCP, a framework that reconciles supervised few-shot adaptation with conformal uncertainty estimation under non-exchangeability. AlignCP learns a reweighted calibration distribution that reduces the score-level discrepancy between the labeled support set and the unlabeled query set. By aligning the one-dimensional nonconformity score distributions, AlignCP aims to close the coverage gap induced by adaptation without requiring query labels.

Xuan Cuong Ngo, Ngan Le

University of Arkansas Fayetteville, Arkansas, USA

VLM Adaptation Conformal Prediction Medical VLMs Adaptive Conformal Inference VLM Calibration

Sep 8, 2026cs.CV

Vision-language models know more about agriculture than they show and rubric-grounded verifications close the gap

Vision-language models (VLMs) show promise for agricultural classification, but zero-shot performance on disease, pest, damage, quality, and species identification remains poor, and it is unclear whether this reflects weak visual features or a failure to connect them to domain knowledge. We build a benchmark of 116 datasets, 834 classes, and 8,324 images spanning these tasks to isolate where the gap arises. Linear probing shows VLM vision encoders already encode agricultural features nearly as separable as a self-supervised DINOv3 baseline, ruling out weak visual representations as the primary bottleneck. Conditioning each model on an oracle reference description (an upper bound on its parametric knowledge) closes most of the gap left by an unaided lower bound, showing VLMs already know more about agriculture than they show. To close this gap without an oracle description at inference time, we structure test-time reasoning around a fixed, per-task diagnostic rubric: the model generates

K

candidate responses and a Probabilistic Pivot Tournament (PPT) verifier, scored pairwise against the rubric, selects the best one. This nearly doubles judged F1 over the lower bound and matches or exceeds the upper bound on several tasks, notably pushing Gemma 4 E4B-it's disease F1 to 0.71, above its own upper bound of 0.60. However, the verifier's letter-scale confidence score has the opposite of its intended effect: filtering to its most confident predictions does not improve accuracy and correlates negatively with correctness across every model and pool size tested, so the score cannot serve as a measure of predictive uncertainty, and most of the observed gain likely comes from rubric-grounded generation rather than pairwise verification.

Earl Ranario, Jared Smith, Lars Lundqvist +2

University of California, Davis

Agricultural Image Analysis Plant Disease Classification VLM Reasoning Classification Rubric-Based Evaluation VLM Calibration

Sep 1, 2026cs.CV

IntroConformal: Conformal Factuality Guarantees for Large Vision-Language Models via Introspective Signals

Large Vision-Language Models (LVLMs) have achieved strong multimodal performance, yet ensuring the factual correctness of generated content remains challenging. Existing methods that provide statistical guarantees on factuality typically rely on external verifiers or generation-time confidence signals, which introduce auxiliary dependencies or often fail for confident but incorrect outputs. We argue that reliable factuality control can instead be achieved through introspective signals derived from the model itself. We introduce IntroConformal, a training-free Conformal Risk Control (CRC) framework that provides finite-sample, distribution-free factuality guarantees. We first instantiate it with layer-wise semantic stability, a conformity score derived from hidden-state representations, and then propose verification probability, a stronger score capturing the model's self-administered judgment on claim factuality. Across multiple LVLM architectures, IntroConformal satisfies the conformal risk guarantee while substantially reducing abstention and achieving competitive or superior claim-level discrimination relative to external verifier-based baselines.

Md. Atabuzzaman, Christian Alexander, Chris Thomas

Department of Computer Science Virginia Tech

VLM Evaluation Large Vision-Language Models VLM Calibration VLM Hallucination Mitigation

Aug 31, 2026cs.LG

Good Memory Has ECC: Evaluating the Memory of Vision-Language Models Beyond Accuracy

Memory is widely viewed as an important unsolved problem for LLMs and VLMs, and current benchmarks typically evaluate it by testing accuracy over long text or video. However, accuracy alone misses properties that matter for real long-horizon tasks. We introduce ECCBench, a benchmark and evaluation protocol that measures memory beyond a system's capacity--its raw accuracy at a specific budget--via three axes we call ECC: efficiency--the computation, in FLOPs, needed to answer from memory; compression--whether compressible inputs are remembered more accurately or efficiently; and calibration--whether the system abstains in response to its own uncertainty and the cost of an error. We find that pretrained VLMs compress their memory over text but not video and are poorly calibrated on both. Among a broader set of memory backbones, several non-Transformer architectures achieve better compression-calibration tradeoffs than RoPE Transformers, suggesting they may be useful components for agents operating over long horizons.

Shmuel Berman, Jia Deng

Department of Computer Science Princeton University

VLM Evaluation Memory-Augmented VLMs VLM Calibration

Aug 31, 2026cs.AI

Rethinking the Test-Time Prompt Tuning Objective from the Perspective of Calibration

Test-time prompt tuning (TPT) has emerged as a powerful paradigm, refining prompts for each test sample via entropy minimization (EM) over multiple augmented views. However, we identify a limitation in the standard EM-based adaptation: it inherently drives the model toward overconfident predictions disregarding sample-specific uncertainty, leading to significant calibration degradation. To address these limitations, we propose a new objective that replaces the conventional EM loss by aligning the original-view prediction with a target distribution derived from augmented views via cross-entropy, while adversarially incorporating the entropy of the target distribution to capture sample-specific uncertainty. Furthermore, to better construct this target distribution, we apply confidence-aware temperature scaling to each augmented-view prediction according to its confidence, sharpening confident predictions while softening uncertain ones. This formulation allows the model to increase confidence only when the target distribution is reliable, while preserving uncertainty when it reflects ambiguous or conflicting augmented-view predictions. Extensive experiments across diverse benchmarks demonstrate that our approach not only achieves state-of-the-art accuracy but also significantly improves model calibration.

Jungwon Choi, Hyeonseo Jang, Kibok Lee +1

Department of Computer Science and Engineering, Chung-Ang University · Department of Statistics and Data Science, Yonsei University

VLM Calibration Uncertainty Calibration Test-Time Prompt Tuning

Aug 31, 2026cs.HC

Frontier vision-language models have overtaken young adults at detecting AI-generated portraits -- but not their calibration

AI image generators now create face portraits that are hard to tell from real photographs. Vision-language models (VLMs) are increasingly proposed to flag such images. We benchmarked 19 VLMs on the same 198 face portraits -- real photographs and identity-matched ChatGPT-4o and Imagen 3 versions -- under the same task as our earlier study of 1,667 adults (85% correct overall; accuracy fell steeply with age). The June-2026 cohort of 14 models only matched adults in their 20s-30s. Four weeks later the ceiling broke. Among five July-2026 releases under the identical protocol, gpt-5.6-sol reached 92.8% balanced accuracy (five-draw mean 92.1%), clearly above adults in their 20s (88.5%), and claude-fable-5 detected every AI image while averaging 91.9%. Model sensitivity now exceeds young adults decisively (d' up to 3.4 versus ~ 2.4). What has not been overtaken is human calibration. Model criteria spread from c = -1.10 to +1.45 while humans sit near zero at every age; both new leaders are biased (+0.44, -0.97), and only a few mid-ranked models approach the human balance. Changing the labelled examples still flipped about one answer in four. The best machines now out-see young adults here, without matching the human balance between suspicion and trust.

Sunwhi Kim, Sunyul Kim, Meounggun Jo +1

Hwasung Medi-Science University, Dept. of Bio-Healthcare · Hwasung Medi-Science University, Dept. of Bio-Healthcare, Republic of Korea · Yonsei University, Graduate School of Engineering, Dept. of Artificial Intelligence +5

VLM Evaluation Vision-Language Models AI-Generated Image Detection VLM Calibration

Aug 30, 2026cs.CV

SpanCalib-VLM: Calibrated Hallucination Span Detection in Vision-Language Models

Detecting hallucinations in Large Vision-Language Models (LVLMs) requires both accurate span localization and well-calibrated confidence scores. Fine-tuned generative VLMs excel at identifying hallucinated text spans but suffer from overconfidence and high inference latency. Discriminative sequence taggers offer deterministic speed and superior calibration but exhibit conservative span recall. We present SpanCalib-VLM, a hybrid dual-system for the SHROOM-Visions Shared Task that combines a multimodal sequence tagger, consisting of XLM-RoBERTa-Large fused with a SigLIP vision encoder via cross-attention, with our fine-tuned generative VLM (Qwen3.5-4B-SHROOM-SFT). Through a Union-Calibrated Fusion strategy, candidate spans from the generative model are re-scored with calibrated probabilities from the sequence tagger. On the SHROOM-Visions English evaluation split, our ensemble achieves a Pearson calibration correlation of 0.41 and an overall IoU of 0.39, with a clean-response IoU of 0.91} and overall detection accuracy of 70.7%. We make our model weights and code publicly available.

Amanuel Gizachew Abebe, Yasmin Moslem

Shaggar Institute of Technology · Shaggar City, Ethiopia · Trinity College Dublin +1

Hallucination Detection in VLMs Hallucination Detection VLM Calibration

Aug 27, 2026cs.CV

MVC-Bench: Benchmarking Calibration of Medical Vision-Language Models

Reliable evaluation of vision-language models (VLMs) and medical vision-language models (Medical-VLMs) requires calibrated confidence, particularly under realistic clinical conditions. However, existing efforts mainly focused on improving accuracy, leaving calibration in the medical domain underexplored. To this end, we propose MVC-Bench, a calibration-centric benchmark for medical image classification with VLMs and Medical-VLMs. MVC-Bench assesses the calibration across three axes: (i) robustness to modality, backbone, and domain shift (ii) effectiveness of calibration strategies and prompt-tuning methods (iii) stability under prompt-template and random-seed variations. The benchmark covers eight different backbones, three medical modalities, including fundus imaging, histopathology, and chest X-ray under in-domain and domain shift settings. It compares post-hoc calibration, train-time calibration, and zero-shot inference methods, together with six prompt-tuning methods. Across more than 1638 controlled experiments, we report accuracy and Expected Calibration Error (ECE) as primary metrics, and further report results with complementary calibration measures, including Maximum Calibration Error (MCE) and Adaptive Calibration Error (ACE). We further investigate the underlying causes of miscalibration in VLMs and Medical-VLMs and propose a simple train-time calibration method, Multi-Class Margin (MCM) regularization, which achieves lowest ECE on 10 out of 12 settings in in-domain and remains competitive under domain shifts. Collectively, MVC-Bench provides a structured evaluation framework and actionable guidance for improving calibration in safety-critical medical workflows.

Ashshak Sharifdeen, Shihab Aaqil Ahamed, Ufaq Khan +6

1Mohamed bin Zayed University of AI · 2Sabaragamuwa University of Sri Lanka · 3Digital Platform Development, SLT PLC

VLM Evaluation VLM Robustness Medical Image Classification Medical VLMs VLM Calibration Medical Image Benchmarks

Aug 13, 2026cs.CV

TRAPSBench: Vision-Language Models Encode but Fail to Express Epistemic Restraint

When visual evidence is occluded or chaotic, models should abstain. In this paper, we show that Vision-Language Models (VLMs) can internally distinguish when abstention is required, but fail to express it anyway. We introduce TRAPSBench, a procedurally generated video benchmark of 1,404 matched physics pairs in which a single targeted change renders the outcome undeterminable from the visual evidence. Furthermore, we introduce Penalized Epistemic Calibration Score (PECS), a new robust metric that requires models to both answer correctly when the outcome is knowable, and abstain when the outcome is not. Across 16 VLMs spanning five families, spontaneous restraint is poor: the best PECS is 0.292. The bottleneck is expression, not perception: linear probes decode answerability from hidden states at up to 0.91 AUROC across physics domains; steering a single-layer void direction causally induces or suppresses abstention. Our results replicate across three open-weight families (Qwen, Gemma, LLaVA). The failure is also more pronounced in visual than textual uncertainty: models detect textual impossibility about 4x more readily than missing visual evidence. Closing this representation--output gap likely requires output-stage interventions.

Fnu Pramono, John Cai, Sourabh Kulkarni

Meta Superintelligence Labs · Reflection AI

VLM Evaluation Vision-Language Models VLM Interpretability VLM Calibration Epistemic Uncertainty

Aug 12, 2026cs.CV

LookBack: Where and How to Score LVLM Responses via Visual Reference Usage

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.

Beomsik Cho, Jinhyeong Kim, Dongseok Lee +1

Yonsei University

VLM Evaluation Hallucination Detection in VLMs Large Vision-Language Models Vision-Language Grounding VLM Calibration Faithfulness in VLMs

Aug 11, 2026cs.CV

When Visual Signals Mislead: A Mechanistic Study of Attribute Hallucination in Vision-Language Models

Attribute hallucination---where vision-language models (VLMs) correctly identify an object but mischaracterize its properties---is prevalent yet mechanistically poorly understood. The dominant explanation, language-prior dominance, has motivated prior-suppression methods, but this explanation has not been directly tested at the attribute level. We present VISOR (Visual-Operational Remediation), a unified framework that couples null-image-based diagnosis with routed remediation. Its VSNR diagnostic decomposes each prediction into a visual logit signal and a language-prior signal. Across 10,791 negative-ground-truth samples from three VLM families and three attribute types, the visual signal strongly predicts false positives, whereas the language-prior signal is near chance. VISOR uses this diagnosis to separate two failure modes: low-margin but directionally correct visual signals in color/state attributes, and low-SNR or misaligned visual signals in material attributes. The same diagnosis routes each query to the appropriate operator: calibration for threshold-placement errors, abstention for training-free low-SNR handling, or targeted visual adaptation for material failures that prior suppression cannot correct. Across Qwen, InternVL, and LLaVA, VISOR reduces attribute false positives without relying on the prior-dominance assumption.

Yufei Zhang, Chenlu Zhan, Hongwei Wang

Zhejiang University

Hallucination Detection in VLMs VLM Adaptation VLM Interpretability VLM Calibration VLM Hallucination Mitigation

Aug 11, 2026cs.CV

CARE: Confidence-Aware Reasoning for Reliable Medical VQA

Reinforcement Fine-Tuning (RFT) has enabled medical Multimodal Large Language Models (MLLMs) to produce Chain-of-Thought (CoT) reasoning for visual question answering, yet these models suffer from

\textit{confidence miscalibration}

---a systematic gap between expressed certainty and actual diagnostic accuracy that undermines clinical trust. We propose

\textbf{CARE}

, a

\textbf{C}

onfidence-

\textbf{A}

ware medical

\textbf{RE}

asoning framework that jointly optimizes accuracy and calibration through a dual-stage pipeline. First, a scalable Medical-CoT synthesis provides structured cold-start data for Supervised Fine-Tuning. Second, Group Relative Policy Optimization (GRPO) with a novel

\textbf{Confidence-Aware Reward (CAR)}

mechanism ties the model's confidence to diagnostic correctness within the reward signal. Across three Medical VQA benchmarks,

\textbf{CARE}

achieves the highest diagnostic accuracy while obtaining the lowest Expected Calibration Error and Hallucination Rate, establishing a foundation for trustworthy clinical decision support. Our code is available at https://github.com/anotherbricki/CARE.

Yuetian Du, Yucheng Wang, Zhenyuan Chen +9

Zhejiang University · Ant Group · University of Michigan +1

Medical VQA Medical VLMs VLM Reasoning VLM Calibration VLM Hallucination Mitigation RL for VLMs

Aug 6, 2026cs.CR

Vision-Language Model Confidence Is Not a Property of the Answer

Vision-language models are increasingly deployed behind a confidence gate: the system reads how confident the model is in its answer and defers when confidence is low. This makes the confidence signal itself worth attacking. We show that a white-box adversary who perturbs only the input image, within an L-infinity budget of 8/255 and while keeping the model's answer byte-identical, can invert the confidence ranking, lowering it on correct answers and raising it on wrong ones until the signal points the wrong way. Most of the inversion persists even when the whole next-token distribution is held near the clean one, so the answer does not determine the confidence attached to it. Confidence is a separate signal read from the same network, and it can be corrupted on its own. A gate reading it is turned against itself, rejecting good answers and accepting wrong ones it was built to catch. Across four vision-language models and three visual question-answering benchmarks, the attack drives the model-internal readouts below chance in 83 of 84 readout-by-cell profiles under an adversary that knows which answers are correct; for the two readouts carrying a disjoint calibration reference, it falls below chance under an adversary that does not. Training a probe on frozen hidden states does not fix this: the robustness it gains is paid for with the information that made it useful. Nor does reading confidence from a separate, independently trained model, which holds up only until the attacker reaches it and then falls into the same regime. How far an answer-preserving adversary can reach a signal governs where it survives; whether a robust and informative readout can be built remains open. For deployment, a gate under this attack can admit most wrong answers it would otherwise catch and, corrected for how often the model is wrong, can leave the system worse off than using no gate at all.

Reza Khanmohammadi, Erfan Miahi, Ivan Brugere +4

Michigan State University · JPMorgan AI Research · Henry Ford Health

VLM Robustness Adversarial Attacks on VLMs VLM Calibration

Aug 6, 2026cs.CL

Confidence Estimation for Financial Vision-Language Models in Chart and Document Understanding

LVLMs are increasingly used to read financial charts, tables, and documents, where a single misread figure can move a decision and the most authoritative-looking answer is sometimes one the model produced without reading the exhibit. The operational question is therefore trust, not accuracy: which answers can be acted on, and which escalated to a reviewer. We evaluate seven confidence estimators, three inference-only and four trained internal probes, across five open-weight LVLMs and four conditions from three financial visual question-answering benchmarks, one bilingual; every probe is trained only on natural images and applied to finance without adaptation, so the results measure out-of-distribution transfer. Three findings hold. First, the scarce property is calibration, not ranking: the inference baselines rank correct above incorrect answers competitively but are badly overconfident, calibration error far above what a threshold can tolerate, and only the trained probes produce a thresholdable score. Second, reliability is structured rather than global, along two axes a practitioner can read directly: the best estimator shifts with both model and task, none leading more than eight of twenty (model, condition) cells, and a controlled bilingual contrast exposes an apparent language robustness as a composition artifact that dissolves once models are read one at a time. Third, cast as deferral under an error budget, how much can be safely automated is set first by the model's competence and only narrowed by its confidence, so deferral clears a real share of the easiest condition and almost none of the hardest, near zero at a strict 5% budget. Two trained probes carry the calibration a deferral policy needs, and among them only the grounding-aware one lowers its confidence on answers a model gives without using the figure, separating detected non-grounding from a fluent guess.

Reza Khanmohammadi, Simerjot Kaur, Charese H. Smiley +2

Michigan State University East Lansing, Michigan, USA · JPMorgan AI Research New York, New York, USA

Visual Question Answering Selective Prediction VLM Calibration Finance

Aug 6, 2026cs.CV

Respect Your Zero-Shot Uncertainty: Conservative Calibration for Test-Time-Adapted Vision-Language Models

Test-time adaptation (TTA) can improve the recognition accuracy of vision-language models under distribution shift, but often degrades calibration, making predictive confidence unreliable for downstream decision-making. Many existing label-free calibration approaches are either coupled to prompt optimization or rely on logit-range statistics that provide only a coarse characterization of the predictive distribution. We show that TTA can increase confidence and reduce entropy even when the top-1 prediction and its correctness remain unchanged, a failure mode we term prediction-preserving sharpening. Across diverse TTA methods and benchmarks, larger entropy reductions relative to paired zero-shot predictions are associated with greater increases in Expected Calibration Error (ECE). On entropy-reduced samples, confidence gains also tend to exceed accuracy gains. Based on these findings, we propose Zero-Shot-Anchored Entropy Calibration (ZAEC), a label-free post-hoc method that uses zero-shot entropy as a sample-specific uncertainty reference. ZAEC selectively restores the zero-shot entropy of sharpened predictions through minimal temperature scaling while leaving all other predictions unchanged. It requires no labeled calibration data or learned parameters and preserves class rankings and classification accuracy. Across five TTA methods and 15 datasets, ZAEC achieves the lowest post-hoc macro-average ECE on ViT-B/16, with consistent gains on RN50.

Jingyan Jiang, Yaru Sun, Xiao Chen +5

1Shenzhen Technology University · 3Shenzhen University · 2Tsinghua University +1

VLM Adaptation Test-Time Adaptation of VLMs Model Calibration VLM Calibration

Aug 4, 2026cs.CV

Perceptual Anchoring: Prototype-Guided Text Calibration for Training-free Open-Vocabulary Semantic Segmentation

Training-free open-vocabulary semantic segmentation (OVSS) partitions an image into semantically distinct regions based on arbitrary text descriptions, without learning any additional parameters. However, existing methods typically focus on improving visual representations while treating text embeddings that encode only generic category concepts as fixed classification references. The resulting semantic gap between these generic concepts and the visual representations that capture the specific appearances of target instances often causes incomplete masks and erroneous predictions in non-target regions. Inspired by the symbol-percept correspondence underlying perceptual anchoring, we propose Prototype-Guided Text Calibration (PTC) for training-free OVSS. In the Perceiving stage, PTC selects reliable visual evidence based on initial matching scores to construct category-specific visual prototypes. In the Anchoring stage, PTC uses these prototypes to calibrate their corresponding text embeddings, with the calibration strength adaptively adjusted based on the amount of visual evidence. Consequently, the calibrated text embeddings align more accurately with instance-specific visual representations while preserving generic category semantics and open-vocabulary generalization. Moreover, PTC requires neither additional training nor external models and can serve as a plug-and-play module for existing methods. Extensive experiments across eight benchmarks show that PTC significantly enhances the performance of six representative methods and yields more complete and accurate segmentation results. These results validate PTC as a simple and effective approach to improving visual-text alignment.

Wanli Ma, Jiangwen Lu, Qinmu Peng +1

School of Electronic Information and Communications, Huazhong University of Science and Technology, Wuhan 430074, China · National Anti-counterfeit Engineering Research Center, Huazhong University of Science and Technology, Wuhan 430074, China · School of Automation, Hangzhou Dianzi University, Hangzhou 310018, China

Open-Vocabulary Semantic Segmentation Vision-Language Alignment VLM Calibration

Aug 3, 2026cs.CV

Confident but Unreliable: A Behavioral Safety Audit of Vision-Language Models on Brain MRI

Vision-language models (VLMs), including medical specialists, are increasingly proposed for medical imaging, yet their stated confidence is rarely evaluated separately from correctness. We use brain MRI as a controlled, high-stakes testbed for a broader failure mode in frontier multimodal systems: models can appear competent while lacking reliable self-knowledge. We present an automatically graded behavioral audit and pilot study of six instruction-tuned VLMs (five general-purpose and one medical specialist) on 4,102 images (4,032 axial/coronal/sagittal MRI slices from 250 subjects plus 70 non-brain/noise controls), with labels derived from public metadata and released expert segmentation masks rather than new human annotation. Across models, answer coverage is near-complete, but verbalized-confidence calibration is poor: ECE ranges from 0.27 to 0.40, mean confidence on incorrect answers ranges from 0.82 to 0.97, and 33-46% of answered items are high-confidence errors. The most accurate model is also the most confident on its errors, while a base/specialist family contrast suggests that medical adaptation improves tumor-presence detection without improving confidence reliability. Open-ended diagnostics further show that hallucination and abstention vary separately from multiple-choice accuracy. These findings argue that medical-image VLM evaluation should report verbalized-confidence reliability, confident error, hallucination, and abstention alongside accuracy.

Amir Sabbaghziarani, Mohammadsajad Abavisani, Sergey Plis

TReNDS Center Atlanta, GA, USA

VLM Evaluation VLM Hallucination Medical VLMs VLM Calibration

Aug 3, 2026cs.AI

Can You Trust the Confidence? ConfBench for Vision-Language Models on Document Extraction

Intelligent document processing (IDP) with vision-language models (VLMs) hinges on confidence scores trustworthy enough to route extractions between automation and human review. Existing document benchmarks are dominated by clean, high-quality samples, leaving low accuracy regions too sparse for calibration assessment. We introduce ConfBench, the first calibration-specific benchmark for key information extraction (KIE), built by applying 20 controlled degradation pipelines to a diverse document set, yielding 1,346 variants and 70K+ entity-level evaluations spanning the full accuracy spectrum. We evaluate four proprietary and three open-weight VLMs under verbalized and log-probability confidence estimation methods across three input modalities, and find: (i) OCR+Image modality results in more accurate confidence estimates; (ii) model capability is the dominant factor: within the Claude family confidence quality scales monotonically with capability, while across families parameter count is a poor predictor; (iii) calibration quality varies widely across models, from near-perfect to severely overconfident, and per-model post-hoc correction rescales these absolute confidence values for threshold-based routing without altering ranking-based operational metrics; and (iv) log-probability with first-token aggregation consistently outperforms mean-token and margin aggregations. We also introduce ECARB, a review-budget metric translating discriminative gains into operational savings. We release ConfBench publicly to enable systematic study of confidence estimators and calibration methods for trustworthy IDP application deployment.

Priyashree Roy, Sujitha Martin, Mohammad Rostami +6

Amazon Web Services

VLM Evaluation Document Understanding Confidence Estimation in Language Models Document Information Extraction VLM Calibration

Jul 31, 2026cs.CV

When Model Priors Conflict with Visual Evidence: Mitigating Commonsense-Driven Hallucinations by Selective Prior Calibration

In vision--language models, commonsense-driven hallucination (CDH) occurs when a model's commonsense prior overrides clear visual evidence of an atypical state. For example, a model may report that a visibly six-fingered hand has five fingers. We show that these errors are systematically directed: when a model answers a question about a counterfactual (CF) image incorrectly, its answer often coincides with the candidate it prefers without access to the image. Suppressing this prior indiscriminately can repair CF errors, but may also disrupt correct answers on matched commonsense (CS) images, where the same prior is helpful. We therefore propose Selective Prior Calibration (SPC), which subtracts candidate-level prior-preference estimates from image-conditioned scores with an instance-dependent strength and revises the original prediction only when the resulting score pattern strongly supports an alternative. Extensive experiments demonstrate that SPC substantially improves accuracy on CF images while largely preserving accuracy on matched CS images. Furthermore, these gains generalize across CDH categories, candidate-answer permutations, and other conflict benchmarks, while SPC rarely alters predictions on benchmarks without such conflicts.

Kesheng Chen, Yamin Hu, Wenjian Luo

Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies Institute of Cyberspace Security School of Computer Science and Technology, Harbin Institute of Technology, Shenzhen, China

VLM Hallucination VLM Calibration VLM Hallucination Mitigation