Vision-Language Models

Also known as VLM

Follow this topic Ask about this topic

Vision-language models study how systems can connect images or videos with language to recognize, describe, retrieve, and reason about visual content. Common approaches include contrastive image-text encoders such as CLIP, generative models that connect vision encoders to language models, and instruction tuning, sometimes combined with retrieval or tools. Researchers evaluate them on image-text retrieval, visual question answering, OCR and document understanding, visual grounding, and video benchmarks. A central challenge is that fluent answers do not guarantee visual grounding: hallucinations, reliance on textual shortcuts, weak compositional and temporal reasoning, and the costs of efficient deployment remain important concerns.

Latest papers

738 papers77 papers in the last four weeks, up 20% on the four weeks before. 0.4% of all new papers.

Mar 7, 2026

VirtueBench: Evaluating Trustworthiness under Uncertainty in Long Video Understanding

Recent Vision-Language Models (VLMs) have made remarkable progress in multimodal understanding tasks, yet their evaluation on long video understanding remains unreliable. Due to limited frame inputs, key frames necessary for answering the question may be missing from the model's input. However, models that truthfully refuse to answer under such uncertainty are marked as incorrect, while those that guess may coincidentally produce the correct answer and thus obtain deceptively higher accuracy, leading to misleading evaluation results and encouraging models to guess rather than respond honestly. To address this issue, we introduce VirtueBench, a benchmark explicitly designed to assess model trustworthiness under uncertainty. VirtueBench constructs multiple frame-sampling levels for each video and provides ground truths that distinguish between answerable and unanswerable cases. Evaluations on 25 open-source and commercial VLMs reveal distinct refusal behaviors across different model families, with refusal accuracy ranging from over 70% in the best models to nearly 0% in the worst. Moreover, most models exhibit a substantial drop in refusal when the prompt does not explicitly require them to do so. These findings highlight the need for developing trustworthy VLMs for multimodal understanding, guided by benchmarks and leaderboards that emphasize reliability and trustworthiness.

Xueqing Yu, Bohan Li, Yan Li +1

Peking University · ByteDance

Abstention in VLMs VLM Evaluation Video-Language Model Evaluation Vision-Language Models Long-Video Understanding

Mar 7, 2026

Looking Back and Forth: Cross-Image Attention Calibration and Attentive Preference Learning for Multi-Image Hallucination Mitigation

Although large vision-language models (LVLMs) have demonstrated remarkable capabilities, they are prone to hallucinations in multi-image tasks. We attribute this issue to limitations in existing attention mechanisms and insufficient cross-image modeling. Inspired by this, we propose a structured hallucination mitigation framework involving Cross-Image Attention calibration and Preference Learning (CAPL). CAPL explicitly enhances inter-image interactions at the architectural level while reinforcing reliance on genuine cross-image evidence during training, thereby improving the model's perception and modeling of cross-image associations. Specifically, we (i) introduce a selectable image token interaction attention mechanism to establish fine-grained cross-image entity alignment and information flow; (ii) design a cross-image modeling-based preference optimization strategy that contrasts reasoning outcomes under full inter-image interaction and those obtained when images are mutually invisible, encouraging the model to ground its predictions in authentic visual evidence and mitigating erroneous inferences driven by textual priors. Experimental results demonstrate that CAPL consistently improves performance across multiple model architectures, achieving stable gains on both multi-image hallucination and general benchmarks. Notably, performance on single-image visual tasks remains stable or slightly improves, indicating strong generalization capability.

Xiaochen Yang, Hao Fang, Jiawei Kong +3

Tsinghua Shenzhen International Graduate School, Tsinghua University · Harbin Institute of Technology · Beijing Institute of Technology +1

VLM Hallucination Mitigation Vision-Language Models VLM Reasoning Cross-Attention

Mar 6, 2026

Penguin-VL: Exploring the Efficiency Limits of VLM with LLM-based Vision Encoders

Vision Language Model (VLM) development has largely relied on scaling model size, which hinders deployment on compute-constrained mobile and edge devices such as smartphones and robots. In this work, we explore the performance limits of compact (e.g., 2B and 8B) VLMs. We challenge the prevailing practice that state-of-the-art VLMs must rely on vision encoders initialized via massive contrastive pretraining (e.g., CLIP/SigLIP). We identify an objective mismatch: contrastive learning, optimized for discrimination, enforces coarse and category-level invariances that suppress fine-grained visual cues needed for dense captioning and complex VLM reasoning. To address this issue, we present Penguin-VL, whose vision encoder is initialized from a text-only LLM. Our experiments reveal that Penguin-Encoder serves as a superior alternative to traditional contrastive pretraining, unlocking a higher degree of visual fidelity and data efficiency for multimodal understanding. Across various image and video benchmarks, Penguin-VL achieves performance comparable to leading VLMs (e.g., Qwen3-VL) in mathematical reasoning and surpasses them in tasks such as document understanding, visual knowledge, and multi-perspective video understanding. Notably, these gains are achieved with a lightweight architecture, demonstrating that improved visual representation rather than model scaling is the primary driver of performance. Our ablations show that Penguin-Encoder consistently outperforms contrastive-pretrained encoders, preserving fine-grained spatial and temporal cues that are critical for dense perception and complex reasoning. This makes it a strong drop-in alternative for compute-efficient VLMs and enables high performance in resource-constrained settings. Code: https://github.com/tencent-ailab/Penguin-VL

Boqiang Zhang, Lei Ke, Ruihan Yang +5

Vision-Language Models Vision-Language Pretraining VLM Reasoning

Mar 6, 2026

GreenRFM: Learning a resource-efficient radiology vision-language foundation model via supervision-centric pre-training

Radiology foundation models (RFMs) have largely inherited the scale-first recipe of natural-image vision--language pre-training. This recipe is difficult to deploy in 3D radiology, where training corpora are smaller, reports vary across institutions, and receiving hospitals often need local adaptation under privacy and compute constraints. We ask whether routine radiology reports can instead be converted into auditable diagnostic supervision that shapes the image encoder, text encoder, aligned space, and local-adaptation procedure. We develop GreenRFM, a supervision-centric pre-training framework organized around four empirical principles: More distilled, Ubiquitous, Semantic-enforcing, and Task-aligning (MUST) supervision. These principles convert noisy reports into structured diagnostic signals and use them to learn discriminative unimodal encoders plus an aligned image--text space for diagnosis-centered multimodal use. GreenRFM requires 24 GPU-hours on a single 24GB GPU (lightweight variant: 6GB VRAM, 4~hours) and reaches a zero-shot CT-RATE AUC of 84.8. Evaluations using more than 200,000 volumes from six institutions and two modalities show transfer to private clinical cohorts and to musculoskeletal MRI. On a local institutional cohort, computationally feasible retraining raises macro-AUC from 70.5 to 82.1. The aligned space also improves hepatocellular-carcinoma microvascular-invasion prediction and trans-arterial chemoembolization response analysis over established clinical scores. These results support supervision-centric pre-training as a practical route to resource-efficient, locally adaptable, diagnosis-centered radiology vision--language representations.

Yingtai Li, Shuai Ming, Qiuli Wang +13

School of Biomedical Engineering, Division of Life Sciences and Medicine, University of Science and Technology of China (USTC), Hefei, Anhui, 230026, China · Center for Medical Imaging, Robotics, Analytic Computing & Learning (MIRACLE), Suzhou Institute for Advanced Research, USTC, Suzhou, Jiangsu, 215123, China · Department of Radiology, The First Affiliated Hospital of USTC, Division of Life Sciences and Medicine, USTC, Hefei, Anhui, 230001, China +7

Medical Imaging Foundation Models Radiology VLMs Vision-Language Models VLM Adaptation Multimodal Pretraining Efficient VLM Training

Mar 6, 2026

Do Foundation Models Know Geometry? Probing Frozen Features for Continuous Physical Measurement

Vision-language models encode continuous geometry that their text pathway fails to express: a 6,000-parameter linear probe extracts hand joint angles at 6.1 degrees MAE from frozen features, while the best text output achieves only 20.0 degrees -- a 3.3x bottleneck. LoRA fine-tuning (r=16, 2,000 images) narrows this gap to 6.5 degrees, providing evidence for a pathway-training deficit rather than a representational one. Training objective determines accuracy more than architecture: five encoders spanning self-supervised, contrastive, and hybrid paradigms converge to statistically equivalent accuracy (R^2 approximately 0.55, TOST-equivalent at delta=0.03) despite sharing as little as CKA=0.41 representational similarity -- functional convergence without representational convergence. Autoregressive generation damages geometric fidelity, but the damage originates in the generation process, not in language alignment: Qwen2.5-VL's LLM layers actually improve probe accuracy over its raw vision encoder. Layer-wise analysis reveals a universal mid-network accuracy peak across all architectures, with attention heads in layers 18-22 carrying disproportionate geometric signal. These findings enable a single frozen backbone to function as a multi-task geometric sensor through lightweight probes, without fine-tuning or text generation.

Yakov Pyotr Shkolnikov

Representation Geometry in VLMs Vision-Language Models Linear Probing

Mar 6, 2026

Do 3D Large Language Models Really Understand 3D Spatial Relationships?

Recent 3D Large-Language Models (3D-LLMs) claim to understand 3D worlds, especially spatial relationships among objects. Yet, we find that simply fine-tuning a language model on text-only question-answer pairs can perform comparably or even surpass these methods on the SQA3D benchmark without using any 3D input. This indicates that the SQA3D benchmark may not be able to detect if the model exploits textual shortcuts rather than engages in 3D-aware reasoning. To address this issue, we introduce Real-3DQA, a more rigorous evaluation benchmark that filters out easy-to-guess questions and introduces a structured taxonomy to assess various aspects of 3D reasoning. Experiments on Real-3DQA confirm that existing 3D-LLMs struggle with spatial relationships once simple cues are removed. We further propose a 3D-reweighted training objective that guides model to rely more on 3D visual clues, substantially enhancing 3D-LLMs performance in spatial reasoning tasks. Our findings underscore the need for robust benchmarks and tailored training strategies to advance genuine 3D vision-language understanding. Project page: https://real-3dqa.github.io/.

Xianzheng Ma, Tao Sun, Shuai Chen +7

VGG, University of Oxford · AVL, University of Oxford · Stanford University +2

3D Spatial Reasoning 3D VQA 3D VLMs Vision-Language Models Spatial Reasoning Benchmarks

Mar 6, 2026

Spatial Colour Mixing Illusions as a Perception Stress Test for Vision-Language Models

Vision-language models (VLMs) achieve strong benchmark results, yet can exhibit systematic perceptual weaknesses: structured, large changes to pixel values can cause confident yet nonsensical predictions, even when the underlying scene remains easily recognizable to humans. We study this gap using Spatial Colour Mixing, a programmatic family of colour distortions that overlays structured patterns (in both RGB and Ostwald colour systems) onto natural images. We introduce a framework of eight spatial colour mixing variants and evaluate nine VLMs across three model families on four datasets. Across models and datasets, accuracy degrades sharply with increasing distortion, and scaling the language model does not reliably mitigate the failure. In a human study with 61 participants on an animal recognition dataset, humans substantially outperform VLMs under the same distortions. Finally, we show that a simple human-inspired preprocessing step recovers a meaningful portion of performance for several distortion types, motivating perception-aware preprocessing and tool-use as practical strategies for improving VLM robustness.

Nicoleta-Nina Basoc, Adrian Cosma, Emilian Radoi

National University of Science and Technology POLITEHNICA Bucharest, Bucharest, Romania · IDSIA, The Dalle Molle Institute for Artificial Intelligence, Lugano, Switzerland

VLM Robustness Vision-Language Models VLM Evaluation

Mar 6, 2026

DeepSight: Bridging Depth Maps and Language with a Depth-Driven Multimodal Model

Multimodal large language models (MLLMs) have achieved impressive performance across various tasks such as image captioning and visual question answer(VQA); however, they often struggle to accurately interpret depth information inherent in visual data. In this work, we introduce DeepSight, the first dedicated depth MLLM designed to enhance three-dimensional scene understanding. Unlike conventional methods that align RGB image encodings with text, our approach takes advantage of the unique characteristics of depth images: single-channel grayscale images where the pixel values directly reflect depth cues to improve spatial reasoning. To address challenges associated with limited depth data and the inadequacy of simple channel replication, we construct a novel depth image-text pair dataset and a depth instruction dataset. Depth maps are generated from visual images using the GLPN model, and GPT-4 is employed to curate corresponding depth instructions, an approach validated by LLaVA. Additionally, we modify the ViT encoder in CLIP to incorporate local object information, thereby capturing the subtle continuous variations of depth more effectively. To evaluate the performance of our model, we develop a comprehensive depth question answer benchmark based on existing depth image datasets, which rigorously assesses understanding in typical depth map scenarios. Experimental results demonstrate that DeepSight significantly enhances depth perception and downstream task performance, marking a substantial step forward in multimodal three-dimensional understanding.

Hao Yang, Hongbo Zhang, Yanyan Zhao +1

Harbin Institute of Technology, Harbin, China

3D VLMs Vision-Language Models 3D Scene Understanding 3D VQA

Mar 6, 2026

Probing Visual Concepts in Lightweight Vision-Language Models for Automated Driving

The use of Vision-Language Models (VLMs) in automated driving applications is becoming increasingly common, with the aim of leveraging their reasoning and generalisation capabilities to handle long-tail scenarios. However, these models often fail on simple visual questions that are highly relevant to automated driving, and the reasons behind these failures remain poorly understood. In this work, we examine the intermediate activations of VLMs and assess the extent to which specific visual concepts are linearly encoded, with the goal of identifying bottlenecks in the flow of visual information. Specifically, we create counterfactual image sets that differ only in a targeted visual concept and then train linear probes to distinguish between them using the activations of five state-of-the-art (SOTA) VLMs, including two training variants for one model. Our results show that concepts such as the presence of an object or agent in a scene are explicitly and linearly encoded, whereas other spatial visual concepts, such as the orientation of an object or agent, are only implicitly encoded by the spatial structure retained by the vision encoder, or are not linearly encoded at all. In parallel, we observe that in certain cases, even when a concept is linearly encoded in the model's activations, the model still fails to answer correctly. This leads us to identify two failure modes. The first is perceptual failure, where the visual information required to answer a question is not linearly encoded in the model's activations. The second is cognitive failure, where the visual information is present but the model fails to align it correctly with language semantics. Finally, we show that increasing the distance of the object in question quickly degrades the linear separability of the corresponding visual concept...

Nikos Theodoridis, Reenu Mohandas, Ganesh Sistu +3

Department of Electronic and Computer Engineering University of Limerick · Valeo Vision Systems

Vision-Language Models VLMs for Autonomous Driving VLM Interpretability

Mar 5, 2026

DARK: Diagonal-Anchored Repulsive Knowledge Distillation for Vision-Language Models under Extreme Compression

Compressing vision-language models for on-device deployment is increasingly important in clinical settings, but knowledge distillation (KD) degrades sharply when the teacher-student capacity gap spans an order of magnitude or more. We argue that, under such gaps, strict imitation of the teacher is a poor objective: much of the teacher's pairwise similarity structure reflects its own architectural biases rather than information a compact student can efficiently represent. We propose Diagonal-Anchored Repulsive Knowledge Distillation (DARK), a contrastive KD framework that decomposes the distillation loss into a diagonal term (matched image-text pairs) and an off-diagonal term (non-target similarities). The diagonal term anchors matched-pair alignment throughout training; the off-diagonal term is annealed from positive to negative weighting, transitioning the student from imitating to repelling the teacher's non-target similarity structure. We instantiate DARK by distilling FetalCLIP, a 427M-parameter fetal ultrasound vision-language model, into MobileFetalCLIP, a 75M-parameter student model with a

26\times

smaller visual encoder, running in 1.6,ms on an iPhone16Pro. The student matches or exceeds its teacher on three zero-shot benchmarks, including HC18 biometry validity (88.6% vs.\ 83.5%) and brain sub-plane F1 (0.784 vs.\ 0.702). Embedding-geometry and logit analyses show that DARK induces structured decorrelation: the student preserves teacher-aligned per-image confidence while diverging from inherited inter-class confusion, suggesting that controlled repulsion can be more efficient than imitation under extreme compression.

Numan Saeed, Asif Hanif, Fadillah Adamsyah Maani +2

Computer Vision, Mohamed Bin Zayed University of Artificial Intelligence, Abu Dhabi, UAE

VLM Distillation Knowledge Distillation Vision-Language Models

Mar 5, 2026

VisionPangu: A Compact and Fine-Grained Multimodal Assistant with 1.7B Parameters

Large Multimodal Models (LMMs) have achieved strong performance in vision-language understanding, yet many existing approaches rely on large-scale architectures and coarse supervision, which limits their ability to generate detailed image captions. In this work, we present VisionPangu, a compact 1.7B-parameter multimodal model designed to improve detailed image captioning through efficient multimodal alignment and high-quality supervision. Our model combines an InternVL-derived vision encoder with the OpenPangu-Embedded language backbone via a lightweight MLP projector and adopts an instruction-tuning pipeline inspired by LLaVA. By incorporating dense human-authored descriptions from the DOCCI dataset, VisionPangu improves semantic coherence and descriptive richness without relying on aggressive model scaling. Experimental results demonstrate that compact multimodal models can achieve competitive performance while producing more structured and detailed captions. The code and model weights will be publicly available at https://www.modelscope.cn/models/asdfgh007/visionpangu.

Jiaxin Fan, Wenpo Song

National Key Laboratory for Novel Software Technology, Nanjing University

Vision-Language Models Multimodal Large Language Models Image Captioning Vision-Language Alignment

Mar 5, 2026

Differentially Private Multimodal In-Context Learning

Vision-language models are increasingly applied to sensitive domains such as medical imaging and personal photographs, yet existing differentially private methods for in-context learning are limited to few-shot, text-only settings because privacy cost scales with the number of tokens processed. We present Differentially Private Multimodal Task Vectors (DP-MTV), the first framework enabling many-shot multimodal in-context learning with formal

(\varepsilon, δ)

-differential privacy by aggregating hundreds of demonstrations into compact task vectors in activation space. DP-MTV partitions private data into disjoint chunks, applies per-layer clipping to bound sensitivity, and adds calibrated noise to the aggregate, requiring only a single noise addition that enables unlimited inference queries. We evaluate on eight benchmarks across three VLM architectures, supporting deployment with or without auxiliary data. At

\varepsilon=1.0

, DP-MTV achieves 50% on VizWiz compared to 55% non-private and 35% zero-shot, preserving most of the gain from in-context learning under meaningful privacy constraints.

Ivoline C. Ngong, Zarreen Reza, Joseph P. Near

University of Vermont, Burlington, VT, USA · Independent Researcher

Privacy Protection in VLMs Multimodal ICL Privacy-Preserving ML Differential Privacy Vision-Language Models

Mar 4, 2026

When Visual Evidence is Ambiguous: Pareidolia as a Diagnostic Probe for Vision Models

When visual evidence is ambiguous, vision models must decide how to interpret face-like patterns. Face pareidolia, the perception of faces in non-face objects, provides a controlled probe of such decisions. We introduce a diagnostic framework that analyzes detection, localization, uncertainty and bias across class, difficulty and emotion. We evaluate six models spanning four representational regimes: vision-language models (VLMs; CLIP-B/32, CLIP-L/14, LLaVA-1.5-7B), pure vision classification (ViT), object detection (YOLOv8), and face detection (RetinaFace). Our results reveal that uncertainty and bias are decoupled: low uncertainty can signal either safe suppression, as in detectors, or extreme over-interpretation, as in VLMs. VLMs exhibit semantic overactivation, systematically interpreting ambiguous non-human regions as Human, with LLaVA over-calling on 73% of non-human pareidolic images, especially for negative emotions. ViT instead follows an uncertainty-as-abstention strategy, remaining diffuse yet largely unbiased. Detection-based models achieve low bias through conservative priors that suppress pareidolia responses even when localization is controlled. Together, these results show that behavior under ambiguity is governed more by representation than thresholds, establishing face pareidolia as a diagnostic of semantic robustness and a source of ambiguity-aware hard negatives for vision models. Code will be released upon publication.

Qianpu Chen, Derya Soydaner, Rob Saunders

LIACS (Leiden Institute of Advanced Computer Science), Leiden University, Leiden, The Netherlands

VLM Robustness VLM Evaluation Vision-Language Models

Mar 4, 2026

From Narrow to Panoramic Vision: Attention-Guided Cold-Start Reshapes Multimodal Reasoning

The cold-start initialization stage plays a pivotal role in training Multimodal Large Reasoning Models (MLRMs), yet its mechanisms remain insufficiently understood. To analyze this stage, we introduce the Visual Attention Score (VAS), an attention-based metric that quantifies how much a model attends to visual tokens. We find that reasoning performance is strongly correlated with VAS (r=0.9616): models with higher VAS achieve substantially stronger multimodal reasoning. Surprisingly, multimodal cold-start fails to elevate VAS, resulting in attention distributions close to the base model, whereas text-only cold-start leads to a clear increase. We term this counter-intuitive phenomenon Lazy Attention Localization. To validate its causal role, we design training-free interventions that directly modulate attention allocation during inference, performance gains of 1

-

2% without any retraining. Building on these insights, we further propose Attention-Guided Visual Anchoring and Reflection (AVAR), a comprehensive cold-start framework that integrates visual-anchored data synthesis, attention-guided objectives, and visual-anchored reward shaping. Applied to Qwen2.5-VL-7B, AVAR achieves an average gain of 7.0% across 7 multimodal reasoning benchmarks. Ablation studies further confirm that each component of AVAR contributes step-wise to the overall gains. The code, data, and models are available at https://github.com/lrlbbzl/Qwen-AVAR.

Ruilin Luo, Chufan Shi, Yizhen Zhang +10

1Tsinghua University · 3Qwen Team, Alibaba Group · University of South California +3

Multimodal Reasoning VLM Reasoning Vision-Language Models Multimodal Instruction Tuning

Mar 3, 2026

Benchmarking Compact VLMs for Clip-Level Surveillance Anomaly Detection Under Weak Supervision

CCTV safety monitoring demands anomaly detectors combine reliable clip-level accuracy with predictable per-clip latency despite weak supervision. This work investigates compact vision-language models (VLMs) as practical detectors for this regime. A unified evaluation protocol standardizes preprocessing, prompting, dataset splits, metrics, and runtime settings to compare parameter-efficiently adapted compact VLMs against training-free VLM pipelines and weakly supervised baselines. Evaluation spans accuracy, precision, recall, F1, ROC-AUC, and average per-clip latency to jointly quantify detection quality and efficiency. With parameter-efficient adaptation, compact VLMs achieve performance on par with, and in several cases exceeding, established approaches while retaining competitive per-clip latency. Adaptation further reduces prompt sensitivity, producing more consistent behavior across prompt regimes under the shared protocol. These results show that parameter-efficient fine-tuning enables compact VLMs to serve as dependable clip-level anomaly detectors, yielding a favorable accuracy-efficiency trade-off within a transparent and consistent experimental setup.

Kirill Borodin, Kirill Kondrashov, Nikita Vasiliev +4

Faculty of Information Technology, Moscow Technical University of Communication and Informatics, Moscow 111024, Russia

Vision-Language Models Weakly Supervised Video Anomaly Detection VLM Adaptation Efficient VLM Inference Parameter-Efficient Fine-Tuning

Mar 3, 2026

Nodes Are Early, Edges Are Late: Probing Diagram Representations in Large Vision-Language Models

Large vision-language models (LVLMs) demonstrate strong performance on diagram understanding benchmarks, yet they still struggle with understanding relationships between elements, particularly those represented by nodes and directed edges (e.g., arrows and lines). To investigate the underlying causes of this limitation, we probe the internal representation of LVLMs using a carefully constructed synthetic diagram dataset based on directed graphs. Our probing experiments reveal that edge information is not linearly separable in the vision encoder and becomes linearly encoded only in the text tokens in the language model. In contrast, node information and global structural features are already linearly encoded in individual hidden states of the vision encoder. These findings suggest that the stage at which linearly separable representations are formed varies depending on the type of visual information. In particular, the delayed emergence of edge representations may help explain why LVLMs struggle with relational understanding, such as interpreting edge directions, which require more abstract, compositionally integrated processes.

Haruto Yoshida, Keito Kudo, Yoichi Aoki +4

Tohoku University · Human Informatics Labs., NTT, Inc.

Vision-Language Models Large Vision-Language Models VLM Reasoning VLM Interpretability Visual Representation Learning

Mar 3, 2026

OmniFashion: Towards Generalist Fashion Intelligence via Multi-Task Vision-Language Learning

Fashion intelligence spans multiple tasks, i.e., retrieval, recommendation, recognition, and dialogue, yet remains hindered by fragmented supervision and incomplete fashion annotations. These limitations jointly restrict the formation of consistent visual-semantic structures, preventing recent vision-language models (VLMs) from serving as a generalist fashion brain that unifies understanding and reasoning across tasks. Therefore, we construct FashionX, a million-scale dataset that exhaustively annotates visible fashion items within an outfit and organizes attributes from global to part-level. Built upon this foundation, we propose OmniFashion, a unified vision-language framework that bridges diverse fashion tasks under a unified fashion dialogue paradigm, enabling both multi-task reasoning and interactive dialogue. Experiments on multi-subtasks and retrieval benchmarks show that OmniFashion achieves strong task-level accuracy and cross-task generalization, highlighting its offering of a scalable path toward universal, dialogue-oriented fashion intelligence.

Zhengwei Yang, Andi Long, Hao Li +3

National Engineering Research Center for Multimedia Software, Institute of Artificial Intelligence, School of Computer Science, Wuhan University · Harbin Institute of Technology

Vision-Language Models Large Vision-Language Models Image Retrieval Recommender Systems

Mar 3, 2026

Mind the Way You Select Negative Texts: Pursuing the Distance Consistency in OOD Detection with VLMs

Out-of-distribution (OOD) detection seeks to identify samples from unknown classes, a critical capability for deploying machine learning models in open-world scenarios. Recent research has demonstrated that Vision-Language Models (VLMs) can effectively leverage their multi-modal representations for OOD detection. However, current methods often incorporate intra-modal distance during OOD detection, such as comparing negative texts with ID labels or comparing test images with image proxies. This design paradigm creates an inherent inconsistency against the inter-modal distance that CLIP-like VLMs are optimized for, potentially leading to suboptimal performance. To address this limitation, we propose InterNeg, a simple yet effective framework that systematically utilizes consistent inter-modal distance enhancement from textual and visual perspectives. From the textual perspective, we devise an inter-modal criterion for selecting negative texts. From the visual perspective, we dynamically identify high-confidence OOD images and invert them into the textual space, generating extra negative text embeddings guided by inter-modal distance. Extensive experiments across multiple benchmarks demonstrate the superiority of our approach. Notably, our InterNeg achieves state-of-the-art performance compared to existing works, with a 3.47% reduction in FPR95 on the large-scale ImageNet benchmark and a 5.50% improvement in AUROC on the challenging Near-OOD benchmark.

Zhikang Xu, Qianqian Xu, Zitai Wang +4

Institute of Information Engineering, Chinese Academy of Sciences · School of Cyber Security, University of Chinese Academy of Sciences · Institute of Computing Technology, Chinese Academy of Sciences +2

OOD Detection Vision-Language Models

Mar 1, 2026

VisNec: Measuring and Leveraging Visual Necessity for Multimodal Instruction Tuning

The effectiveness of multimodal instruction tuning depends not only on dataset scale, but critically on whether training samples genuinely require visual reasoning. However, existing instruction datasets often contain a substantial portion of visually redundant samples (solvable from text alone), as well as multimodally misaligned supervision that can degrade learning. To address this, we propose VisNec (Visual Necessity Score), a principled data selection framework that measures the marginal contribution of visual input during instruction tuning. By comparing predictive loss with and without visual context, VisNec identifies whether a training instance is vision-critical, redundant, or misaligned. To preserve task diversity, we combine VisNec with semantic clustering and select high-necessity samples within each cluster. Across 10 downstream benchmarks, training on only 15% of the LLaVA-665K dataset selected by VisNec achieves 100.2% of full-data performance. On the smaller Vision-Flan-186K dataset, our selection not only further reduces data size but also surpasses full-data training by 15.8%. These results demonstrate that measuring and leveraging visual necessity provides an effective solution for both efficient and robust multimodal instruction tuning. Codes and selected subsets will be released upon acceptance.

Mingkang Dong, Hongyi Cai, Jie Li +4

SJTU · Universiti Malaya · CMU +3

Vision-Language Models Data Selection Training Data Selection Multimodal Instruction Tuning Visual Instruction Tuning

Mar 1, 2026

GameVerse: Can Vision-Language Models Learn from Video-based Reflection?

Human gameplay is a visually grounded interaction loop in which players act, reflect on failures, and watch tutorials to refine strategies. Can Vision-Language Models (VLMs) also learn from video-based reflection? We present GameVerse, a comprehensive video game benchmark that enables a reflective visual interaction loop. Moving beyond traditional fire-and-forget evaluations, it uses a novel reflect-and-retry paradigm to assess how VLMs internalize visual experience and improve policies. To facilitate systematic and scalable evaluation, we also introduce a cognitive hierarchical taxonomy spanning 15 globally popular games, dual action space for both semantic and GUI control, and milestone evaluation using advanced VLMs to quantify progress. Our experiments show that VLMs benefit from video-based reflection in varied settings, and perform best by combining failure trajectories and expert tutorials-a training-free analogue to reinforcement learning (RL) plus supervised fine-tuning (SFT).Our project page is available at https://gameverse-bench.github.io/ . Our code is available at https://github.com/THUSI-Lab/GameVerse .

Kuan Zhang, Dongchen Liu, Qiyue Zhao +5

College of AI, Tsinghua University, China.

Vision-Language Models VLM Evaluation

Mar 1, 2026

ClinCoT: Clinical-Aware Visual Chain-of-Thought for Medical Vision Language Models

Medical Vision-Language Models have shown promising potential in clinical decision support, yet they remain prone to factual hallucinations due to insufficient grounding in localized pathological evidence. Existing medical alignment methods primarily operate at the response level through preference optimization, improving output correctness but leaving intermediate reasoning weakly connected to visual regions. Although chain-of-thought (CoT) enhances multimodal reasoning, it remains largely text-centric, limiting effective integration of clinical visual cues. To address this gap, we propose ClinCoT, a clinical-aware visual chain-of-thought framework that transforms preference optimization from response-level correction to visual-driven reasoning. We introduce an automatic data generation pipeline that constructs clinically grounded preference pairs through reasoning with hypotheses-driven region proposals. Multiple Med-LLMs evaluators rank and assign scores to each response, and these rankings serve as supervision to train the target model. We further introduce a scoring-based margin-aware optimization strategy that incorporates both preference ranking and score difference to refine region-level reasoning trajectories. To maintain alignment as the model's policy evolves during training, we adopt an iterative learning scheme that dynamically regenerates preference data. Extensive experiments on three medical VQA and report generation benchmarks demonstrate that ClinCoT consistently improves factual grounding and achieves superior performance compared with existing preference-based alignment methods.

Xiwei Liu, Yulong Li, Xinlin Zhuang +5

Mohamed bin Zayed University of Artificial Intelligence · East China Normal University

Vision-Language Models Medical VLMs VLM Hallucination Mitigation Medical VQA

Mar 1, 2026

Unified Vision-Language Modeling via Concept Space Alignment

We introduce V-SONAR, a vision-language embedding space extended from the text-only embedding space SONAR (Omnilingual Embeddings Team et al., 2026), which supports 1500 text languages and 177 speech languages. To construct V-SONAR, we propose a post-hoc alignment pipeline that maps the representations of an existing vision encoder into the SONAR space. We thoroughly evaluate V-SONAR and show that its embeddings achieve competitive performance on text-to-video retrieval. Equipped with the OMNISONAR text decoder, V-SONAR further surpasses state-of-the-art vision-language models on video captioning tasks, including DREAM-1K (BLEU 23.9 vs. 19.6) and PE-VIDEO (BLEU 39.0 vs. 30.0). Leveraging V-SONAR, we first demonstrate that the Large Concept Model (LCM; LCM team et al. 2024) operating in SONAR and trained with English text only, can perform both single- and multi-visual concept understanding in a zero-shot manner. Finally, we introduce V-LCM, which extends the LCM with vision-language instruction tuning. V-LCM encodes vision and language inputs into an unified sequence of latent embeddings via V-SONAR and SONAR, and it is trained with the same latent diffusion objective for next-embedding prediction as in LCM's text-only pre-training. Experiments on a large-scale multilingual and -modal instruction-tuning data mixture highlight the potential of V-LCM: V-LCM matches state-of-the-art vision-language models on tasks covering image/video captioning and question answering, while significantly outperforming them across 61 rich- to low-resource languages out of all 62 tested languages.

Yifu Qiu, Paul-Ambroise Duquenne, Holger Schwenk

University of Edinburgh · FAIR at Meta

Vision-Language Models Vision-Language Alignment Video-Language Models

Feb 28, 2026

TokenCom: Vision-Language Model for Multimodal and Multitask Token Communications

Visual-Language Models (VLMs), with their strong capabilities in image and text understanding, offer a solid foundation for intelligent communications. However, their effectiveness is constrained by limited token granularity, overlong visual token sequences, and inadequate cross-modal alignment. To overcome these challenges, we propose TaiChi, a novel VLM framework designed for token communications. TaiChi adopts a dual-visual tokenizer architecture that processes both high- and low-resolution images to collaboratively capture pixel-level details and global conceptual features. A Bilateral Attention Network (BAN) is introduced to intelligently fuse multi-scale visual tokens, thereby enhancing visual understanding and producing compact visual tokens. In addition, a Kolmogorov Arnold Network (KAN)-based modality projector with learnable activation functions is employed to achieve precise nonlinear alignment from visual features to the text semantic space, thus minimizing information loss. Finally, TaiChi is integrated into a multimodal and multitask token communication system equipped with a joint VLM-channel coding scheme. Experimental results validate the superior performance of TaiChi, as well as the feasibility and effectiveness of the TaiChi-driven token communication system.

Feibo Jiang, Siwei Tu, Li Dong +5

Hunan Provincial Key Laboratory of Intelligent Computing and Language Information Processing, Hunan Normal University, Changsha, China. · School of Information Science and Engineering, Hunan Normal University, Changsha, China. · Changsha Social Laboratory of Artificial Intelligence, Hunan University of Technology and Business, Changsha, China. +5

Joint Source-Channel Coding Vision-Language Models Visual Token Compression Vision-Language Alignment

Feb 28, 2026

U-VLM: Hierarchical Vision Language Modeling for Report Generation

Automated radiology report generation is key for reducing radiologist workload and improving diagnostic consistency, yet generating accurate reports for 3D medical imaging remains challenging. Existing vision-language models face two limitations: they do not leverage segmentation-pretrained encoders, and they inject visual features only at the input layer of language models, losing multi-scale information. We propose U-VLM, which enables hierarchical vision-language modeling in both training and architecture: (1) progressive training from segmentation to classification to report generation, and (2) multi-layer visual injection that routes U-Net encoder features to corresponding language model layers. Each training stage can leverage different datasets without unified annotations. U-VLM achieves state-of-the-art performance on CT-RATE (F1: 0.414 vs 0.258, BLEU-mean: 0.349 vs 0.305) and AbdomenAtlas 3.0 (F1: 0.624 vs 0.518 for segmentation-based detection) using only a 0.1B decoder trained from scratch, demonstrating that well-designed vision encoder pretraining outweighs the benefits of 7B+ pre-trained language models. Ablation studies show that progressive pretraining significantly improves F1, while multi-layer injection improves BLEU-mean. Code is available at https://github.com/yinghemedical/U-VLM.

Pengcheng Shi, Minghui Zhang, Kehan Song +3

Medical Image Insights Co. Ltd., Shanghai, China · Shanghai Jiao Tong University, Shanghai, China

Radiology Report Generation Vision-Language Models CT Report Generation 3D VLMs

Feb 28, 2026

Self-Correction Inside the Model: Leveraging Layer Attention to Mitigate Hallucinations in Large Vision Language Models

Although Large Vision-Language Models (LVLMs) have made substantial progress, hallucination, where generated text is not grounded in the visual input, remains a challenge. As LVLMs become stronger, previously reported hallucination patterns, such as linguistic bias and overthinking phenomenon, become far less consistent, making the corresponding mitigation techniques substantially less effective. In this paper, we introduce an Internal self-Correction mechanism utilizing Layer Attention (ICLA) that operates directly on hidden states during generation. Each layer selectively retrieves information from all preceding layers through a diagonal cross-layer attention mechanism, enabling self-refinement without any external correction signals. With introducing and training only 0.2M and 0.1M additional parameters on LLaVA1.5-7B and Qwen2.5-VL-7B, \ours consistently improves visual grounding across multiple hallucination benchmarks, demonstrating its effectiveness for more advanced LVLMs.

April Fu

Independent Researcher

Cross-Layer Attention Self-Correction in VLMs VLM Hallucination Mitigation Vision-Language Models Large Vision-Language Models Transformer Attention

Feb 27, 2026

CC-VQA: Conflict- and Correlation-Aware Method for Mitigating Knowledge Conflict in Knowledge-Based Visual Question Answering

Knowledge-based visual question answering (KB-VQA) demonstrates significant potential for handling knowledge-intensive tasks. However, conflicts arise between static parametric knowledge in vision language models (VLMs) and dynamically retrieved information due to the static model knowledge from pre-training. The outputs either ignore retrieved contexts or exhibit inconsistent integration with parametric knowledge, posing substantial challenges for KB-VQA. Current knowledge conflict mitigation methods primarily adapted from language-based approaches, focusing on context-level conflicts through engineered prompting strategies or context-aware decoding mechanisms. However, these methods neglect the critical role of visual information in conflicts and suffer from redundant retrieved contexts, which impair accurate conflict identification and effective mitigation. To address these limitations, we propose CC-VQA: a novel training-free, conflict- and correlation-aware method for KB-VQA. Our method comprises two core components: (1) Vision-Centric Contextual Conflict Reasoning, which performs visual-semantic conflict analysis across internal and external knowledge contexts; and (2) Correlation-Guided Encoding and Decoding, featuring positional encoding compression for low-correlation statements and adaptive decoding using correlation-weighted conflict scoring. Extensive evaluations on E-VQA, InfoSeek, and OK-VQA benchmarks demonstrate that CC-VQA achieves state-of-the-art performance, yielding absolute accuracy improvements of 3.3% to 6.4% compared to existing methods. Code is available at https://github.com/cqu-student/CC-VQA.

Yuyang Hong, Jiaqi Gu, Yujin Lou +7

School of Artificial Intelligence, UCAS · MAIS, Institute of Automation · Alibaba Cloud Computing

Knowledge-Based VQA Knowledge Conflicts in Language Models VLM Reasoning Vision-Language Models

Feb 27, 2026

Half-Truths Break Similarity-Based Retrieval

When a text description is extended with an additional detail, image-text similarity should drop if that detail is wrong. We show that CLIP-style dual encoders often violate this intuition: appending a plausible but incorrect object or relation to an otherwise correct description can increase the similarity score. We call such cases half-truths. On COCO, CLIP prefers the correct shorter description only 34.0% of the time, and performance drops to 27.8% when the added detail is a relation. We trace this vulnerability to weak supervision on caption parts: contrastive training aligns full sentences but does not explicitly enforce that individual entities and relations are grounded. We propose CS-CLIP (Component-Supervised CLIP), which decomposes captions into entity and relation units, constructs a minimally edited foil for each unit, and fine-tunes the model to score the correct unit above its foil while preserving standard dual-encoder inference. CS-CLIP raises half-truth accuracy to 76.4% and improves average performance on established compositional benchmarks by 5.8 points, suggesting that reducing half-truth errors aligns with broader gains in compositional understanding.

Bora Kargi, Arnas Uselis, Seong Joon Oh

University of Tübingen, Tübingen AI Center · Konrad Zuse School of Excellence in Learning and Intelligent Systems (ELIZA)

Contrastive Learning for VLMs Vision-Language Models Vision-Language Retrieval Vision-Language Alignment Compositional Representation Learning Counterfactual Evaluation of VLMs

Feb 26, 2026

Cytoarchitecture in Words: Weakly Supervised Vision-Language Modeling for Human Brain Microscopy

Vision foundation models increasingly support interactive scientific workflows, but natural-language interaction requires coupling visual representations to language. Curated image-text pairs for this coupling are scarce in many biomedical domains. This is the case for cell-body-stained histological sections of the human brain, where microscopic image patches encode cytoarchitecture: cellular density, morphology, laminar structure, and areal organization. We propose retrieve-and-enrich supervision, a weakly supervised scheme for training image-conditioned language models without curated image-text pairs. The method retrieves label-level text from the literature via a shared anatomical label, then enriches it with image-specific properties such as cortical layer thickness and cell density. Labels provide training targets only, not model inputs. We use this scheme to couple CytoNet, a cytoarchitectonic vision foundation model, to an open-weight large language model via a lightweight Flamingo-style adapter. Across 57 brain areas, the resulting model produces plausible cytoarchitectonic descriptions, supports open-set use by rejecting out-of-scope areas, and predicts the correct area for in-scope patches with 90.6% accuracy. Removing explicit area names from generated text still leaves descriptions sufficient for a language model to recover the area in an 8-way test with 68.6% accuracy. A second instruction-tuned model trained with image-specific targets recovers cortical layer thickness and cell-densities from individual patches, allowing the model to read out information beyond canonical area descriptions. Our results show that retrieve-and-enrich supervision offers a practical route to vision-language training in specialized imaging domains where labels and expert text exist but image-level captions do not.

Matthew Sutton, Katrin Amunts, Timo Dickscheid +1

Institute of Neuroscience and Medicine (INM-1), Research Centre Jülich, Jülich, Germany · Helmholtz AI, Research Centre Jülich, Jülich, Germany · C´ecile & Oskar Vogt Institute for Brain Research, University Hospital Düsseldorf, Düsseldorf, Germany +1

Medical VLMs Vision-Language Models

Feb 26, 2026

Where Vision Becomes Text: Locating the OCR Routing Bottleneck in Vision-Language Models

Vision-language models (VLMs) can read text from images, but where does this optical character recognition (OCR) information enter the language processing stream? We investigate the OCR routing mechanism across three architecture families (Qwen3-VL, Phi-4, InternVL3.5) using causal interventions. By computing activation differences between original images and text-inpainted versions, we identify architecture-specific OCR bottlenecks whose dominant location depends on the vision-language integration strategy: DeepStack models (Qwen) show peak sensitivity at mid-depth (about 50%) for scene text, while single-stage projection models (Phi-4, InternVL) peak at early layers (6-25%), though the exact layer of maximum effect varies across datasets. The OCR signal is remarkably low-dimensional: PC1 captures up to 72.9% of variance. Crucially, principal component analysis (PCA) directions learned on one dataset transfer to others, demonstrating shared text-processing pathways. Surprisingly, in models with modular OCR circuits (notably Qwen3-VL-4B), OCR removal can improve counting performance (up to +6.9 percentage points), suggesting OCR interferes with other visual processing in sufficiently modular architectures.

Jonathan Steinberg, Oren Gal

Swarms & AI Lab (SAIL), University of Haifa

Vision-Language Models VLM Interpretability

Feb 24, 2026

Exploring Vision-Language Models for Open-Vocabulary Zero-Shot Action Segmentation

Temporal Action Segmentation (TAS) requires dividing videos into action segments, yet the vast space of activities and alternative breakdowns makes collecting comprehensive datasets infeasible. Existing methods remain limited to closed vocabularies and fixed label sets. In this work, we explore the largely unexplored problem of Open-Vocabulary Zero-Shot Temporal Action Segmentation (OVTAS) by leveraging the strong zero-shot capabilities of Vision-Language Models (VLMs). We introduce a training-free pipeline that follows a segmentation-by-classification design: Frame-Action Embedding Similarity (FAES) matches video frames to candidate action labels, and Similarity-Matrix Temporal Segmentation (SMTS) enforces temporal consistency. Beyond proposing OVTAS, we present a systematic study across 14 diverse VLMs, providing the first broad analysis of their suitability for open-vocabulary action segmentation. Experiments on standard benchmarks show that OVTAS achieves strong results without task-specific supervision, underscoring the potential of VLMs for structured temporal understanding.

Asim Unmesh, Kaki Ramesh, Mayank Patel +2

Purdue University, USA · Birla Institute of Technology and Science (BITS) Hyderabad, India

Temporal Action Segmentation Zero-Shot Learning Vision-Language Models