Vision-Language Models

Also known as VLM

Follow this topic Ask about this topic

Momentum

55 papers in the last four weeks, up 120% on the four weeks before. 0.5% of all new papers.

Jul 13Week of Sep 28

Latest papers 491

May 12, 2026cs.CV

VIP: Visual-guided Prompt Evolution for Efficient Dense Vision-Language Inference

Pursuing training-free open-vocabulary semantic segmentation in an efficient and generalizable manner remains challenging due to the deep-seated spatial bias in CLIP. To overcome the limitations of existing solutions, this work moves beyond the CLIP-based paradigm and harnesses the recent spatially-aware dino

.

txt framework to facilitate more efficient and high-quality dense prediction. While dino

.

txt exhibits robust spatial awareness, we find that the semantic ambiguity of text queries gives rise to severe mismatch within its dense cross-modal interactions. To address this, we introduce Visual-guided Prompt evolution (VIP) to rectify the semantic expressiveness of text queries in dino

.

txt, unleashing its potential for fine-grained object perception. Towards this end, VIP integrates alias expansion with a visual-guided distillation mechanism to mine valuable semantic cues, which are robustly aggregated in a saliency-aware manner to yield a high-fidelity prediction. Extensive evaluations demonstrate that VIP: 1. surpasses the top-leading methods by 1.4%-8.4% average mIoU, 2. generalizes well to diverse challenging domains, and 3. requires marginal inference time and memory overhead.

Hao Zhu, Shuo Jin, Wenbin Liao +4

Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China · School of Computer Science and Technology, University of Chinese Academy of Sciences, Beijing, China · XJTLU +1

Image Segmentation Vision-Language Models Efficient VLM Inference Prompt Optimization Open-Vocabulary Semantic Segmentation Vision-Language Alignment

May 12, 2026cs.CV

Beyond Text Prompts: Visual-to-Visual Generation as A Unified Paradigm

Humans often specify and create through visual artifacts: typography sheets, sketches, reference images, and annotated scenes. Yet modern visual generators still ask users to serialize this intent into text, a bottleneck that compresses signals like spatial structure, exact appearance, and glyph shape. We propose \textbf{\emph{visual-to-visual} (V2V)} generation, in which the user conditions a generative model with a visual specification page rather than a text prompt. The page is not an edit target, but a visual document that specifies the desired output. We introduce \textbf{V2V-Zero}, a training-free framework that exposes this interface in existing vision-language model (VLM) conditioned generators by replacing text-only conditioning with final-layer hidden states extracted from visual pages, exploiting the fact that the frozen VLM already maps both text and images into the generator's conditioning space. On GenEval, V2V-Zero reaches 0.85 with a frozen Qwen-Image backbone, closely matching its optimized text-to-image performance without fine-tuning. To evaluate the broader V2V space, we introduce \textbf{Simple-V2V Bench}, spanning seven visual-conditioning tasks and seven models, including GPT Image 2, Nano Banana 2, Seedream 5.0 Lite, open-weight baselines, and a video extension. V2V-Zero scores 32.7/100, outperforming evaluated open-weight image baselines and revealing a clear capability hierarchy: attribute binding is strong, content generation is unreliable, and structural control remains hard even for commercial systems. A HunyuanVideo-1.5 extension scores 20.2/100, showing the interface transfers beyond images. Mechanistic analysis shows the default reasoning path is primarily visually routed, with 95.0% of conditioning-token attention mass on visual-page hidden states.

Yaofang Liu, Kangning Cui, Meng Chu +7

City University of Hong Kong · City University of Hong Kong (Dongguan) · The Hong Kong University of Science and Technology +4

Vision-Language Models Conditional Image Generation Generative AI Evaluation

May 12, 2026cs.CV

UHR-Micro: Diagnosing and Mitigating the Resolution Illusion in Earth Observation VLMs

Vision-Language Models (VLMs) are increasingly used to analyze ultra-high-resolution (UHR) Earth observation imagery, yet they face a severe scale mismatch between broad scene context and micro-scale targets. We refer to this phenomenon as a "resolution illusion": higher input resolution provides access to more visual detail, but does not necessarily translate into reliable perception of task-relevant micro-evidence. To benchmark this challenge, we introduce UHR-Micro, a benchmark comprising 11,072 instructions grounded in 1,212 UHR images, designed to evaluate VLMs on micro-scale evidence in native-resolution Earth observation imagery. UHR-Micro spans diverse target scales, task families, and visual conditions, with each sample having an objectively verifiable target. Experiments with representative high-resolution VLMs show substantial failures in localizing and interpreting task-relevant evidence, despite access to high-resolution inputs. Further analysis shows that model scaling alone is insufficient, while localized evidence substantially improves performance, pointing to evidence access as a major bottleneck. Motivated by this finding, we propose Micro-evidence Active Perception (MAP), which constructs a spatially grounded evidence state from localized observations and supplements it when the current evidence is insufficient. Across two backbone VLMs, MAP improves UHR-Micro performance by 8.65 percentage points on average. UHR-Micro and MAP provide a framework for diagnosing and improving high-resolution reasoning in Earth observation VLMs. Datasets and source code were released at https://github.com/MiliLab/UHR-Micro.

Shuo Ni, Tong Wang, Jing Zhang +5

National Key Laboratory of Science and Technology on Space-Born Intelligent Information Processing, Beijing Institute of Technology, Beijing, China · Zhongguancun Academy, Beijing, China · School of Computer Science, Wuhan University, Wuhan, China +2

VLM Evaluation Remote Sensing Image Understanding Vision-Language Models VLM Reasoning Visually Grounded Reasoning Remote Sensing Benchmarks+1

May 12, 2026cs.CV

DistractMIA: Black-Box Membership Inference on Vision-Language Models via Semantic Distraction

Vision-language models (VLMs) are trained on large-scale image-text corpora that may contain private, copyrighted, or otherwise sensitive data, motivating membership inference as a tool for training-data auditing. This is especially challenging for deployed VLMs, where auditors typically observe only generated textual responses. Existing VLM membership inference attacks either rely on probability-level signals unavailable in such settings, or use mask-based semantic prediction tasks whose effectiveness depends on object-centric visual assumptions. To address these limitations, we propose DistractMIA, an output-only black-box framework based on semantic distraction. Rather than removing visual evidence, DistractMIA preserves the original image, inserts a known semantic distractor, and measures how generated responses change. This design is motivated by the intuition that member samples remain more anchored to the original image semantics, while non-member samples are more easily redirected toward the distractor. To make this signal reliable, DistractMIA calibrates distractor configurations on a reference set and derives membership scores from repeated textual generations, capturing response stability and distractor uptake without accessing logits, probabilities, or hidden states. Experiments across multiple VLMs and benchmarks show that DistractMIA consistently outperforms both output-only and stronger-access baselines. Its performance on a medical benchmark further demonstrates applicability beyond object-centric natural images.

Hongyi Tang, Zhihao Zhu, Yi Yang

The Hong Kong University of Science and Technology

Vision-Language Models Membership Inference Attacks Privacy Leakage in Language Models

May 12, 2026cs.CV

Urban Risk-Aware Navigation via VQA-Based Event Maps for People with Low Vision

Visual impairment affects hundreds of millions of people worldwide, severely limiting their ability to navigate urban environments safely and independently. While wearable assistive devices offer a promising platform for real-time hazard detection, existing approaches rely on task-specific vision pipelines that lack flexibility and generalizability. In this work, we propose an event map framework based on visual question answering that leverages Vision-Language Models (VLMs) for pedestrian scene description and hazard identification across diverse real-world environments, using a three-level hierarchical query structure to enable fine-grained scene understanding without task-specific retraining. Model responses are aggregated into a weighted risk scoring system that maps street segments into four discrete safety categories, producing navigable risk-aware event maps for route planning. To support evaluation and future research, we introduce a geographically diverse dataset spanning 20 cities across six continents, comprising over 800 annotated images and 18,000 answered questions. We benchmark four VQA architectures -ViLT, LLaVA, InstructBLIP, and Qwen-VL- and find that generative Multimodal Large Language Models (MLLMs) substantially outperform classification-based approaches, with Qwen-VL achieving the best overall balance of precision and recall. These results demonstrate the viability of MLLMs as a flexible and generalizable foundation for assistive navigation systems for visually impaired people.

Antoni Valls, Jordi Sanchez-Riera

Institut de Rob`otica i Inform`atica Industrial, CSIC-UPC, Llorens i Artigas 4-6, 08028 Barcelona, Spain

Visual Question Answering Vision-Language Models

May 12, 2026cs.CV

Revisiting Shadow Detection from a Vision-Language Perspective

Shadow detection is commonly formulated as a vision-driven dense prediction problem, where models rely primarily on pixel-wise visual supervision to distinguish shadows from non-shadow regions. However, this formulation can become unreliable in visually ambiguous cases, where similar dark regions may correspond either to cast shadows or to intrinsically dark surfaces, making visual evidence alone insufficient for establishing a stable decision rule. In this work, we revisit shadow detection from a vision--language perspective and argue that robust prediction benefits from an explicit semantic reference beyond visual cues alone. We propose SVL, a Shadow Vision--Language framework that uses language as an explicit semantic reference to disambiguate shadows from visually similar dark regions. SVL aligns global image representations with shadow-related text embeddings through scene-level shadow ratio regression, and transfers this semantic guidance to dense prediction via global-to-local coupling and local patch-level constraints. Built on a frozen DINOv3 image encoder, SVL learns only lightweight projection and decoding modules, yielding a parameter-efficient design with less than

1\%

trainable parameters. Extensive experiments on multiple shadow detection benchmarks, including dedicated hard-case evaluations, suggest strong overall performance and improved robustness under visually ambiguous conditions. Code is available at https://github.com/harrytea/SVL.

Yonghui Wang, Shaokai Liu, Wengang Zhou +2

Department of Electronic Engineering and Information Science, University of Science and Technology of China · School of Management, Hefei University of Technology · School of Artificial Intelligence and Data Science, University of Science and Technology of China

Vision-Language Models

May 11, 2026cs.AI

Don't Look at the Numbers: Visual Anchoring Bias and Layer-wise Representation in VLMs

Embedded numeric anchors on images systematically bias Vision-Language Model quality judgments across six VLMs from five architectural families (ANOVA eta^2 = 0.18-0.77, all p < 0.001). Anchor effects are 2.5x larger than severe image quality degradation, confirming bias is not reducible to visual changes. Layer-wise probing reveals consistent dissociation: layers where anchor classification saturates (L12-L34) are suboptimal for quality prediction, with optimal layers deeper (R^2 = 0.69-0.91). Fusion analysis identifies architecture-dependent integration -- instant fusion at L1-L2 in two models versus partial or no fusion in three others. These results establish a causal account of visual anchoring bias, linking behavioral susceptibility to representation dynamics.

M. Shalankin

VLM Evaluation Vision-Language Models Representation Learning VLM Interpretability

May 11, 2026cs.CV

Birds of a Feather Flock Together: Background-Invariant Representations via Linear Structure in VLMs

Vision-language models (VLMs), such as CLIP and SigLIP 2, are widely used for image classification, yet their vision encoders remain vulnerable to systematic biases that undermine robustness. In particular, correlations between foreground objects and their backgrounds constitute a salient and practically important class of spurious dependencies. In this work, we revisit the well-known property of high linear additivity in VLM embedding spaces and show that it enables a decomposition of scene representations into foreground and background components. Leveraging this insight, we introduce a pre-training approach that exploits this property to construct background-invariant representations using synthetic data. Our method achieves, to our knowledge, the first worst-group accuracy exceeding

90\%

on Waterbirds under perfect (

100\%

) spurious correlation (i.e., no minority-group examples in the training data). Furthermore, it demonstrates strong sim-to-real transfer and requires no access to real-world debiased data, making it practical for real-world deployment.

Youssef Zaazou, Mark Thomas

Independent Researcher

VLM Robustness Vision-Language Models Spurious Correlation Robustness Invariant Representation Learning Vision-Language Pretraining

May 11, 2026cs.CV

Personal Visual Context Learning in Large Multimodal Models

As wearable devices like smart glasses integrate Large Multimodal Models (LMMs) into the continuous first-person visual streams of individual users, the evolution of these models into true personal assistants hinges on visual personalization: the ability to reason over visual information unique to the wearer. We formalize this capability as Personal Visual Context Learning (Personal VCL), the prompt-time capability of using user-specific visual context to resolve personalized queries. To systematically evaluate this, we present Personal-VCL-Bench, a comprehensive benchmark capturing the personal visual world across persons, objects, and behaviors. Our analysis of frontier LMMs identifies a profound context utilization gap, revealing that the mechanisms for leveraging visual evidence, as well as aggregating multiple visual observations, remain critically understudied. Motivated by these findings, we propose the Agentic Context Bank, a strong inference-time baseline that structures a user's visual context into a self-refining memory bank and employs query-adaptive evidence selection. Our baseline approach consistently improves over standard context prompting regimes across tasks and evaluated backbones, demonstrating a practical path towards future personalized LMMs.

Zihui Xue, Ami Baid, Sangho Kim +2

The University of Texas at Austin

Vision-Language Models Multimodal Large Language Models LLM Context Management Memory-Augmented Language Model Agents

May 11, 2026cs.CL

Grounded or Guessing? LVLM Confidence Estimation via Blind-Image Contrastive Ranking

Large vision-language models suffer from visual ungroundedness: they can produce a fluent, confident, and even correct response driven entirely by language priors, with the image contributing nothing to the prediction. Existing confidence estimation methods cannot detect this, as they observe model behavior under normal inference with no mechanism to determine whether a prediction was shaped by the image or by text alone. We introduce BICR (Blind-Image Contrastive Ranking), a model-agnostic confidence estimation framework that makes this contrast explicit during training by extracting hidden states from a frozen LVLM twice: once with the real image-question pair, and once with the image blacked out while the question is held fixed. A lightweight probe is trained on the real-image hidden state and regularized by a ranking loss that penalizes higher confidence on the blacked-out view, teaching it to treat visual grounding as a signal of reliability at zero additional inference cost. Evaluated across five modern LVLMs and seven baselines on a benchmark covering visual question answering, object hallucination detection, medical imaging, and financial document understanding, BICR achieves the best cross-LVLM average on both calibration and discrimination simultaneously, with statistically significant discrimination gains robust to cluster-aware analysis at 4-18x fewer parameters than the strongest probing baseline.

Reza Khanmohammadi, Erfan Miahi, Simerjot Kaur +4

Michigan State University · Independent AI Researcher · JPMorgan AI Research +1

VLM Evaluation Vision-Language Models Confidence Estimation in Language Models Vision-Language Grounding Visually Grounded Reasoning VLM Calibration

May 11, 2026cs.CV

Verification Mirage: Mapping the Reliability Boundary of Self-Verification in Medical VQA

Self-verification, re-invoking the same vision language model (VLM) in a fresh context to check its own generated answer, is increasingly used as a default safety layer for medical visual question answering (VQA). We argue that this practice is fundamentally unreliable. We introduce [METHOD NAME], a diagnostic framework for mapping the reliability boundary of medical VLM self-verification by decomposing verifier behavior into discrimination capability and agreement bias. Because the verifier and answer generator are capacity-coupled, the verifier can overly agree with the generator, creating a verification mirage: a regime with both high verifier error and high agreement bias, driven by false acceptance of incorrect answers. Evaluating six open-weight VLMs across five medical VQA datasets and seven medical tasks, we find that this boundary is strongly task-conditioned. Knowledge-intensive clinical tasks fall deepest into the mirage, simpler tasks are more resistant, and perceptual tasks lie in between. Verification also fails to provide an independent safety signal: logistic mixed-effects analysis shows that verifier error and agreement bias become more likely when the generator is wrong, while saliency analyses show that verifiers under-attend to image evidence relative to generators, a phenomenon we call the lazy verifier. Cross-verification reduces but does not eliminate the mirage. Moreover, when verification is reused in multi-turn actor-verifier loops, most initially wrong answers become locked in by false verification. Since our experiments use clean benchmarks, the observed reliability boundary likely underestimates failures in real clinical deployment.

Ruinan Jin, Beidi Zhao, Myeongkyun Kang +2

1The University of British Columbia · 2Vector Institute · 3Redmin University of China

VLM Evaluation Vision-Language Models LLM Answer Verification Healthcare Medical VQA Medical VLMs

May 11, 2026cs.CV

LLaVA-CKD: Bottom-Up Cascaded Knowledge Distillation for Vision-Language Models

Large Vision-Language Models (VLMs) are successful in addressing a multitude of vision-language understanding tasks, such as Visual Question Answering (VQA), but their memory and compute requirements remain a concern for practical deployment. A promising class of techniques for mitigating this concern is Knowledge Distillation, where knowledge from a high-capacity Teacher network is transferred to a considerably smaller Student network. However, the capacity gap between the two networks is both a blessing and a curse: the smaller the Student network, the better its efficiency, and the larger the Teacher, the more knowledge it carries; yet, beyond a point, the larger capacity gap between the two leads to worse knowledge transfer. To counter this effect, we propose a bottom-up cascaded knowledge distillation (CKD) framework. Instead of treating knowledge transfer as an activity involving one high-capacity Teacher (or an ensemble of such), inspired by human formal education systems, we introduce one (potentially, more) additional Teacher(s) of intermediate capacity that gradually bring the Student network to the next level, where the next (higher-capacity) Teacher can take over. We provide a theoretical analysis in order to study the effect of cascaded distillation in the generalization performance of the Student. We apply the proposed framework on models build upon the LLaVA methodology and evaluate the derived models on seven standard, publicly available VQA benchmarks, demonstrating their SotA performance.

Nikolaos Gkalelis, Vasileios Mezaris

CERTH-ITI, Thessaloniki, Greece, 57001

Vision-Language Models Efficient VLM Inference VLM Distillation Multi-Teacher Knowledge Distillation Knowledge Distillation

May 11, 2026cs.CV

MFVLR: Multi-domain Fine-grained Vision-Language Reconstruction for Generalizable Diffusion Face Forgery Detection and Localization

The swift advancement in photo-realistic face generation technology has sparked considerable concerns across society and academia, emphasizing the requirement of generalizable face forgery detection and localization methods. Prior works tend to capture face forgery patterns across multiple domains using image modality, other modalities like fine-grained texts are not comprehensively investigated, which restricts the generalization capability of models. Besides, they usually analyze facial images created by GAN, but struggle to identify and localize those synthesized by diffusion. To solve the problems, in this paper, we devise a novel multi-domain fine-grained vision-language reconstruction (MFVLR) model, which explores comprehensive and diverse visual forgery traces via language-guided face forgery representation learning, to achieve generalizable diffusion-synthesized face forgery detection and localization (DFFDL). Specifically, we devise a fine-grained language transformer that studies general fine-grained language embeddings using language reconstruction. We propose a multi-domain vision encoder to capture general and complementary visual forgery patterns across the image and residual domains. A vision decoder is designed to reconstruct image appearance and achieve forgery localization. Besides, we propose an innovative plug-and-play vision injection module to enhance the interaction between the vision and language embeddings. Extensive experiments and visualizations demonstrate that our network outperforms the state of the art on different settings like cross-generator, cross-forgery, and cross-dataset evaluations.

Yaning Zhang, Tianyi Wang, Zan Gao +3

Faculty of Computer Science and Technology, Qilu University of Technology (Shandong Academy of Sciences), Jinan, 250353, China · School of Computing, National University of Singapore, 21 Lower Kent Ridge Rd, 11907750, Singapore · Shandong Artificial Intelligence Institute, Qilu University of Technology (Shandong Academy of Sciences), Jinan, 250014, China +2

Vision-Language Models Image Forgery Detection Image Forgery Localization Deepfake Detection AI-Generated Image Detection

May 11, 2026cs.CV

The Cartesian Shortcut: Re-evaluate Vision Reasoning in Polar Coordinate Space

As current Multimodal Large Language Models rapidly saturate canonical visual reasoning benchmarks, a key question emerges: do these strong scores genuinely reflect robust visual understanding? We identify a pervasive vulnerability, the Cartesian Shortcut: models frequently discretize the orthogonal grid-based layouts prevalent in visual reasoning benchmarks into explicit textual coordinates, offloading reasoning from visual perception to text-based deduction. To re-evaluate visual reasoning when this shortcut is unavailable, we introduce Polaris-Bench, which re-formulates 53 visual reasoning tasks in Polar coordinate space with paired Cartesian counterparts that preserve task semantics, disrupting the orthogonal structure that models exploit. Comprehensive evaluation across

14

state-of-the-art MLLMs reveals that frontier models achieving

69

--

83\%

on Cartesian layouts collapse to

31

--

39\%

on Polar equivalents. Moreover, thinking gains largely vanish on Polar layouts, prompting interventions fail to close the gap, and comparable drops arise on other non-orthogonal layouts. These findings reveal that current MLLMs' visual reasoning performance is strongly coupled to orthogonal grid structure, a fragility that is consistent across model families but far smaller in humans, who maintain 88.8% accuracy on Polar layouts. The benchmark, evaluation suite, and leaderboard are publicly available at https://google-deepmind.github.io/polaris-bench.

Xia Hu, Zhenrui Yue, Brian Potetz +4

Google DeepMind · Stanford University · Google Research

VLM Evaluation Spatial Reasoning Benchmarks Vision-Language Models Visual Reasoning Visually Grounded Reasoning Multimodal Foundation Models+1

May 10, 2026cs.CV

Reflection Anchors for Propagation-Aware Visual Retention in Long-Chain Multimodal Reasoning

Long chain-of-thought (CoT) reasoning improves large vision--language models, but visual information often fades during generation, limiting long-horizon multimodal reasoning. Existing methods either re-inject vision at inference or train policies for stronger grounding, but where to intervene relies on perception heuristics rather than principled gain analysis, and how local visual influence propagates remains implicit. We study this problem from an information-theoretic standpoint and derive a lower bound on the downstream visual gain of a one-step intervention, which suggests two factors: local branching room (token entropy) and downstream visual propagation potential (suffix divergence from a vision-marginalized reference). Guided by this analysis, we propose reflection-anchor policy optimization (RAPO), a GRPO-based policy optimization method that selects high-entropy reflection anchors and optimizes a chain-masked finite-window KL surrogate for downstream visual dependence. Experiments on reasoning-intensive and general-domain benchmarks show that RAPO delivers substantial gains over strong baselines across multiple LVLM backbones. Mechanism analyses further indicate that reflection anchors are enriched for visually sensitive decision points and that RAPO increases contrastive visual-dependence signals along generated trajectories.

Xuan Gong, Hanbo Huang, Hao Zheng +4

Shanghai Jiao Tong University · Lanzhou University

Vision-Language Models Vision-Language Grounding VLM Reasoning Multimodal CoT Reasoning Visually Grounded Reasoning RL for VLMs

May 9, 2026cs.CV

KEPIL: Knowledge-Enhanced Prompt-Image Learning for Prompt-Robust Disease Detection

Vision--language models (VLMs) show promise for clinical decision support in radiology because they enable joint reasoning over radiological images and clinical text, thereby leveraging complementary clinical information. However, radiological findings are long-tailed in practice, leaving some conditions underrepresented and making zero-shot inference essential. Yet current CLIP-style medical VLMs are sensitive to prompt variations and often lack trustworthy external knowledge at inference time, which hinders reliable clinical deployment. We present \textit{KEPIL}, a prompt-robust framework that integrates curated medical knowledge to stabilize zero-shot generalization. KEPIL comprises: (i) \emph{dynamic prompt enrichment} using ontologies with LLM assistance, (ii) a \emph{semantic-aware contrastive loss} aligning embeddings of equivalent prompt variants via a dual-embedding objective, and (iii) \emph{entity-centric report standardization} to yield ontology-aligned representations. Across seven benchmarks, KEPIL achieves state-of-the-art zero-shot inference performance; under prompt-variation tests, it improves AUC by

6.37\%

on \textit{CheXpert} and by

4.11\%

on average. These results suggest that structured knowledge and robust prompt design are key to clinically reliable radiology-facing VLMs. Code will be released at https://github.com/Roypic/KEPIL.

Haozhe Luo, Shelley Zixin Shu, Ziyu Zhou +2

ARTORG Center for Biomedical Engineering Research, University of Bern · Kaiko.AI, Zurich, Switzerland · Shanghai Jiao Tong University +1

VLM Robustness Radiology VLMs Vision-Language Models Healthcare Medical Image Classification Zero-Shot Image Classification

May 9, 2026cs.CV

Gate-and-Merge: Zero-shot Compositional Personalization of Vision Language Models

This paper tackles compositional personalization of vision-language models (VLMs). In this problem, multiple user-defined concepts must be recognized or described jointly at test time. We introduce Gate-and-Merge, a zero-shot framework that enables compositional personalization without the need for co-occurrence training. During personalization, each concept is learned independently as a lightweight LoRA adapter, paired with a concept token. The base model remains unchanged and concepts are kept disentangled. At inference, we enable composition by merging concept-specific LoRA updates directly in weight space. To suppress irrelevant activations and prevent interference, a gating mechanism is employed to estimate textual and visual cues and select only the modules that contribute to the prediction. We further stabilize composition by combining only the most meaningful and mutually consistent updates, helping preserve each concept's identity. Our quantitative and qualitative analyses show consistent gains in performance across multiple personalization tasks in both single-concept and compositional settings.

Guodong Ding, Angela Yao

National University of Singapore

Vision-Language Models VLM Adaptation Model Merging

May 8, 2026cs.CV

ZAYA1-VL-8B Technical Report

We present ZAYA1-VL-8B, a compact mixture-of-experts vision-language model built upon our in-house language model, ZAYA1-8B. Despite its compact size, ZAYA1-VL achieves performance competitive with leading base models such as Molmo2-4B and InternVL3.5-4B, while surpassing models including Qwen2.5-VL-3B, PLM-3B, and MolmoE-1B across a range of image understanding, reasoning, and counting benchmarks. The architecture incorporates two key innovations: (1) vision-specific LoRA adapters integrated into the LLM to increase modality-specific capacity without increasing the number of experts, and (2) bidirectional attention over image tokens within the LLM to enhance visual understanding. We detail the full training pipeline including data composition at each stage, sequence packing, and the attention masking scheme. The model comprises 9.2B total parameters, with 1.4B active parameters including the vision encoder, and is publicly available at https://huggingface.co/Zyphra/ZAYA1-VL.

Hassan Shapourian, Kasra Hejazi, Olabode M. Sule +1

Zyphra Technologies Inc, San Francisco, CA

Vision-Language Models Mixture-of-Experts Language Models VLM Reasoning Multimodal Foundation Models

May 8, 2026cs.CV

PolarVLM: Bridging the Semantic-Physical Gap in Vision-Language Models

Mainstream vision-language models (VLMs) fundamentally struggle with severe optical ambiguities, such as reflections and transparent objects, due to the inherent limitations of standard RGB inputs. While polarization imaging captures polarimetric physical parameters that resolve these ambiguities, existing methods are constrained by fixed-format outputs and remain isolated from open-ended reasoning. To bridge this semantic-physical gap, we introduce PolarVLM, the first multimodal framework integrating polarimetric physical parameters into VLMs. By employing a dual-stream architecture and a progressive two-stage training strategy, PolarVLM effectively prevents physical misinterpretations while preserving general visual abilities. Complementing our architecture, we construct PolarVQA, the first benchmark for polarization-aware VQA, featuring 75K physics-grounded instruction-tuning pairs targeting reflective and transparent scenes. Experiments show that PolarVLM surpasses the RGB baseline by 25.4% overall across five evaluation tasks, with remarkable gains of 26.6% in reflection recognition and 34.0% in glass counting, successfully unlocking physics-aware semantic understanding.

Yuliang Li, Chu Zhou, Heng Guo +3

Beijing University of Posts and Telecommunications, China · National Institute of Informatics, Japan · Peking University, China +1

Visual Question Answering Vision-Language Models Polarimetric Imaging

May 8, 2026cs.CV

Tracing the Arrow of Time: Diagnosing Temporal Information Flow in Video-LLMs

The Arrow-of-Time (AoT) task, determining whether a video plays forward or backward by recognizing temporal irreversibility, is one humans solve with near-perfect accuracy, yet frontier Video Large Language Models (Video-LLMs) perform only modestly above chance. This gap raises a key question: do visual backbones fail to encode temporal information, or does information bottleneck lie elsewhere in the Video-LLM architecture? We address this question by isolating the vision encoder from the Video-LLM and tracing temporal information across the encoder, projector, and LLM. We find that video-centric encoders with explicit temporal modeling encode strong temporal signals, whereas frame-centric encoders do not. However, when video-centric representations are passed through a standard Video-LLM architecture, performance often collapses, revealing a bottleneck of temporal information flow. We identify projector design as a key factor: Q-Former disrupts temporal information, while a time-preserved MLP projection substantially improves the LLM's access to such information. Our layer-wise analysis further shows temporal representation dynamics across encoder layers. Guided by these findings, we build a Video-LLM with temporal-aware video-centric encoder, time-preserved projector, and AoT supervision, surpassing human performance on AoT

_{PPB}

with 98.1% accuracy, and improving broader temporal reasoning tasks by up to 6.0 points on VITATECS-Direction and 1.3 points on TVBench. Our results show that temporal reasoning in Video-LLMs requires both effective temporal encoding and reliable transfer of this information to the LLM.

Peitao Han, Fei Cheng, Lis K. Pereira +2

The University of Osaka · Center for Information and Neural Networks · National Institute of Information and Communications Technology +2

Vision-Language Models Temporal Reasoning in Language Models Video-Language Models Video Reasoning Temporal Video Understanding

May 8, 2026cs.AI

From Pixels to Prompts: Vision-Language Models

When you read a paper about a new Vision-Language Model today, it can be easy to forget how strange this idea would have sounded not so long ago. Teaching machines to see was already hard. Teaching them to read and generate language was already hard. Asking them to do both at once - and then to reason, answer questions, follow instructions, and sometimes even surprise us - still carries a quiet trace of science fiction, even as it becomes routine. This book was born from a simple feeling: it is too easy to get lost. The field moves quickly, new model names appear constantly, and the gap between "I know the buzzwords" and "I actually understand how this works" can feel uncomfortably wide. I have felt that gap many times. If you are holding this book, you probably have too. My goal is not to provide an exhaustive catalog of every dataset, benchmark, and new model variant. Instead, I want to offer something more modest - and, I hope, more durable: a clear mental map of Vision-Language Models. Enough structure that you can read new papers with confidence; enough intuition that you can design your own systems without feeling as if you are assembling LEGO bricks blindly.

Khang Hoang Nhat Vo

Vision-Language Models Large Vision-Language Models VLM Reasoning

May 8, 2026cs.CV

Hierarchical Dual-Subspace Decoupling for Continual Learning in Vision-Language Models

Class-incremental learning aims to continuously acquire new knowledge while preserving previously learned information, thereby mitigating catastrophic forgetting. Existing methods primarily restrict parameter updates but often overlook their structural properties in high-dimensional spaces. From a subspace perspective, updates induced by different tasks tend to lie in multiple overlapping low-rank subspaces, leading to cross-task subspace interference and severe forgetting. To address this issue, we propose HDSD, a Hierarchical Dual-Subspace Decoupling framework for continual learning in vision-language models. Specifically, we introduce a lightweight Feature Modulation Module (FMM) that explicitly decomposes the parameter space into general and task-specific subspaces. Building on this design, we develop two complementary components. First, a General Fusion Module (GFM) evaluates relative parameter changes across tasks and uses an adaptive threshold to capture stable and transferable knowledge. Second, a Hierarchical Learning Module (HLM) performs structured parameter decomposition via Singular Value Decomposition (SVD) and uses a scaling mechanism to constrain updates within distinct subspace scales. Together, these designs reduce subspace interference and parameter drift. Extensive experiments on conventional benchmarks show that HDSD achieves state-of-the-art results.

Mengxin Qin, Xiang Zhang, Kun Wei +2

School of Electronic Engineering, Xidian University Xi’an 710071, China

Class-Incremental Learning Vision-Language Models VLM Adaptation Continual Learning Incremental Learning Continual Learning for VLMs

May 8, 2026cs.CV

DIMoE-Adapters: Dynamic Expert Evolution for Continual Learning in Vision-Language Models

Continual learning enables vision-language models to accumulate knowledge and adapt to evolving tasks without retraining from scratch. However, in multi-domain task-incremental learning, large domain shifts intensify the stability-plasticity dilemma. Most existing methods rely on fixed architectures with statically allocated parameters, which limits adaptation to new domains and aggravates catastrophic forgetting. To address these challenges, we propose DIMoE-Adapters, a Dynamic Incremental Mixture-of-Experts Adapters framework that introduces a dynamic expert evolution paradigm to balance stability and plasticity. This paradigm is implemented through two collaborative components: Self-Calibrated Expert Evolution (SCEE) and Prototype-Guided Expert Selection (PGES). SCEE constructs and evolves a sparse expert pool through expert optimization dynamics, improving plasticity while reducing redundant capacity. PGES controls expert utilization based on the pool shaped by SCEE, improving stability across both previously encountered and unseen tasks. Extensive experiments show that DIMoE-Adapters outperforms previous state-of-the-art methods across various settings.

Mengxin Qin, Xiang Zhang, Xi Wang +3

School of Electronic Engineering, Xidian University Xi’an 710071, China

Vision-Language Models Fine-Tuning VLM Adaptation Mixture of Experts Adapter Tuning Incremental Learning+1

May 8, 2026cs.CV

ForgeVLA: Federated Vision-Language-Action Learning without Language Annotations

Vision-Language-Action (VLA) models hold great promise for general-purpose robotic intelligence, yet scaling up such models is severely bottlenecked by the high cost of acquiring annotated training data. Fortunately, vision-equipped robots deployed across various domains already produce abundant vision-action pairs that can be leveraged to scale up VLA training more efficiently. However, these raw data cannot be centrally aggregated due to various constraints and also exhibit severe heterogeneity. To address these challenges, in this paper, we propose ForgeVLA, a federated VLA training framework that learns VLA models from distributed vision-action pairs without centralizing raw data or requiring manual annotations. Specifically, each client in ForgeVLA is equipped with an embodied instruction classifier that maps vision-action pairs to a predefined instruction set, recovering the missing language modality and forming complete vision-language-action triplets. Beyond triplet construction, we also identify vision-language feature collapse as a critical challenge that has been largely overlooked in prior federated VLA research. To mitigate this issue, ForgeVLA combines a client-side contrastive planning loss with a server-side adaptive aggregation strategy to learn task-discriminative representations efficiently. Extensive experiments across multiple benchmarks show that ForgeVLA significantly outperforms other baselines, and ablation studies further validate the contribution of each component.

Yuhao Zhou, Yunpeng Zhu, Yang Zhou +7

Sichuan University · Engineering Research Center of Machine Learning and Industry Intelligence, Ministry of Education · Zhejiang University +2

Vision-Language Models Efficient VLA Models Vision-Language-Action Models VLMs for Robotics Federated Learning

May 8, 2026cs.CV

GPO-V: Jailbreak Diffusion Vision Language Model by Global Probability Optimization

Diffusion Vision-Language Models (dVLMs), built upon the non-causal foundations of Diffusion Large Language Models (dLLMs), have demonstrated remarkable efficacy in multimodal tasks by departing from the traditional autoregressive generation paradigm. While dVLMs appear inherently robust against conventional jailbreak tactics, which we categorize as Fixed Prefix Optimization (FPO) (e.g., anchoring responses with "Sure, here is"), this perceived resilience is deceptive. Our investigation into the safety landscape of dVLMs reveals a unique refusal pattern: Immediate Refusal and Progressive Refusal. We find that while FPO-based attacks often fail by triggering the latter, the progressive refinement process itself uncovers a novel, latent attack surface. To exploit this vulnerability, we propose Global Probability Optimization (GPO), a general jailbreak paradigm designed specifically for the denoising trajectory of masked diffusion models. Unlike prefix-based methods, GPO manipulates the global generative dynamics to bypass guardrails in diffusion language models. Building on this, we introduce GPO-V, the first visual-modality jailbreak framework tailored for dVLMs. Empirical results demonstrate that GPO-V produces stealthy perturbations with exceptional cross-model transferability, revealing a critical security gap in non-sequential generative architectures. Our findings underscore the critical urgency of addressing safety alignment in dVLMs. These results necessitate an immediate and fundamental re-evaluation of current defense paradigms to mitigate the unique risks of diffusion-based generation. Our code is available at: https://anonymous.4open.science/r/GPO-V-0250.

Yu Pan, Andi Zhang, Yi Wang +2

ShanghaiTech University · University of Warwick · SUN YAT-SEN UNIVERSITY

Diffusion Model Alignment Adversarial Attacks on VLMs Vision-Language Models Diffusion Language Models

May 7, 2026cs.CV

GeoStack: A Framework for Quasi-Abelian Knowledge Composition in VLMs

We address the challenge of knowledge composition in Vision-Language Models (VLMs), where accumulating expertise across multiple domains or tasks typically leads to catastrophic forgetting. We introduce GeoStack (Geometric Stacking), a modular framework that allows independently trained domain experts to be composed into a unified model. By imposing geometric and structural constraints on the adapter manifold, GeoStack ensures the foundational knowledge of the base model is preserved. Furthermore, we mathematically demonstrate a weight-folding property that achieves constant-time inference complexity (

O(1)

), regardless of the number of integrated experts. Experimental results across multi-domain adaptation and class-incremental learning show that GeoStack provides an efficient mechanism for long-term knowledge composition while significantly mitigating catastrophic forgetting. Code is available at https://github.com/QuantitativeImagingLaboratory/GeoStack.

Pranav Mantini, Shishir K. Shah

Department of Computer Science University of Houston · The University of Oklahoma School of Computer Science

Vision-Language Models VLM Adaptation Efficient VLM Inference Continual Learning Catastrophic Forgetting

May 7, 2026cs.CV

PlotPick: AI-powered batch extraction of numerical data from scientific figures

Systematic reviews and meta-analyses often need numerical data reported only in figures, and extracting them with interactive digitisers usually requires a person to select and calibrate each figure. We present PlotPick, an open-source tool that uses vision-language models (VLMs) to extract tabular data from batches of scientific figures, and we benchmark the kind of model it calls: nine VLMs from four providers on ChartX and six of them on PlotQA, against DePlot, a dedicated chart-to-table model, with every system scored on the same items by numeric F1 (F1 over unlabelled numbers at 5% relative tolerance). On six ChartX chart types (n=300) all nine VLMs outperform DePlot in aggregate, at 79.1-96.0% against 74.3%. The lead comes mainly from box plots, where DePlot scores 24.8% against 64.2-97.3%; pooled over the other five types, seven VLMs keep a lead of 4.7 to 11.6 points and the two weakest do not. On a subset of the PlotQA test split (n=529; 427 horizontal bar charts), scored by a lenient best-series variant of the metric, DePlot reaches 87.0%; the two strongest VLMs are level with it or slightly above it, and four fall 3.8 to 30.3 points below. DePlot was trained on PlotQA's training split. Both benchmarks use synthetic charts, and the metric ignores which series a value belongs to. The application itself was not evaluated: its figure detection, structured output, prompt and default model were not tested, and the one benchmarked model it offers, Claude Haiku 4.5, is one of the four below DePlot on PlotQA. Accuracy on biomedical figures has not been established, and every extracted value must be checked against its source figure. This version corrects version 1, which scored most PlotQA replies against category labels instead of plotted values; its claim that every VLM outperformed DePlot on both benchmarks is withdrawn. PlotPick is available at https://plotpick.streamlit.app/.

Tommy Carstensen

Copenhagen Research Centre for Biological and Precision Psychiatry, Mental Health Centre Copenhagen, Copenhagen University Hospital, Copenhagen, Denmark

Data Visualization Vision-Language Models Document Information Extraction Chart QA Chart Data Extraction Chart Understanding

May 7, 2026cs.CV

When Language Overwrites Vision: Over-Alignment and Geometric Debiasing in Vision-Language Models

Vision-Language Models (VLMs) increasingly power high-stakes applications, from medical imaging to autonomous systems, yet they routinely hallucinate, confidently describing content not present in the input. We investigate the root causes of these failure modes with a mechanistic analysis focusing on the decoder-based VLMs. We trace these failure modes to a geometric over-alignment: to bridge the modality gap required by attention mechanisms, decoder-based VLMs over-align visual embeddings with the text manifold, injecting a statistical linguistic bias that systematically overshadows fine-grained visual evidence. While prior work either aggressively closes this gap or suppresses hallucinations through expensive black-box decoding strategies, none addresses the underlying geometric cause. We provide the first quantitative characterization of this over-alignment, demonstrating that linguistic bias concentrates in the top principal components of a universal, dataset-agnostic text subspace. Building on this insight, we propose two complementary remedies: a training-free inference strategy and a bias-aware fine-tuning paradigm, both of which explicitly project out this subspace from visual representations. Our methods significantly reduce hallucinations across POPE, CHAIR, and AMBER benchmarks, and improve CLAIR scores on long-form captioning tasks, with the training-free variant adding no computational overhead over the base model.

Harshvardhan Saini, Samyak Jha, Yiming Tang +1

Indian Institute of Technology Dhanbad · National University of Singapore

Vision-Language Models Object Hallucination in VLMs VLM Interpretability Modality Gap in VLMs Vision-Language Alignment Faithfulness in VLMs+1

May 7, 2026cs.LG

VisMMOE: Exploiting Visual-Expert Affinity for Efficient Visual-Language MoE Offloading

Large-scale vision-language mixture-of-experts (VL-MoE) models provide strong multimodal capability, but efficient deployment on memory-constrained platforms remains difficult. Existing MoE offloading systems are largely designed for text-centric workloads and become much less effective for visual-heavy inputs, where large numbers of visual tokens induce broader and less predictable expert accesses. We present VisMMoE, a VL-MoE offloading system built on a single systems insight: pruning redundant visual tokens can improve offloading not only by reducing computation, but also by reshaping expert demand. We refer to this effect as \textit{visual-expert affinity}: token pruning makes expert accesses more concentrated within layers and more stable across layers, producing a smaller and more predictable expert working set. Guided by this insight, VisMMoE combines affinity-aware token compression, lookahead expert prediction, and cache/pipeline orchestration to improve expert locality and prefetch effectiveness under tight memory budgets. We implement VisMMoE on multiple frameworks and evaluate it on representative VL-MoE models and benchmarks. VisMMoE improves end-to-end inference performance by up to 2.68x and 1.61x, respectively, over strong baselines for today's VL-MoE deployments while maintaining competitive accuracy.

Cheng Xu, Xiaofeng Hou, Jiacheng Liu +1

Shanghai Jiao Tong University Minhang Qu, Shanghai, China

Expert Offloading Vision-Language Models Efficient VLM Inference Mixture-of-Experts Inference Visual Token Compression

May 7, 2026cs.AI

Large Vision-Language Models Get Lost in Attention

Despite the rapid evolution of training paradigms, the decoder backbone of large vision--language models (LVLMs) remains fundamentally rooted in the residual-connection Transformer architecture. Therefore, deciphering the distinct roles of internal modules is critical for understanding model mechanics and guiding architectural optimization. While prior statistical approaches have provided valuable attribution-based insights, they often lack a unified theoretical basis. To bridge this gap, we propose a unified framework grounded in information theory and geometry to quantify the geometric and entropic nature of residual updates. Applying this unified framework reveals a fundamental functional decoupling: Attention acts as a subspace-preserving operator focused on reconfiguration, whereas FFNs serve as subspace-expanding operators driving semantic innovation. Strikingly, further experiments demonstrate that replacing learned attention weights with predefined values (e.g., Gaussian noise) yields comparable or even superior performance across a majority of datasets relative to vanilla models. These results expose severe misallocation and redundancy in current mechanisms, suggesting that state-of-the-art LVLMs effectively ``get lost in attention'' rather than efficiently leveraging visual context.

Gongli Xi, Ye Tian, Mengyu Yang +5

School of Cyberspace Security, Beijing University of Posts and Telecommunications, Beijing, China · State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, Beijing, China · School of Computer Science (National Pilot Software Engineering School), Beijing University of Posts and Telecommunications, Beijing, China +3

Transformer FFNs Vision-Language Models Transformer Self-Attention Large Vision-Language Models VLM Interpretability+1