Large Vision Language Models

End-to-end (E2E) autonomous driving aims to learn a direct mapping from visual observations to control actions. However, these E2E models often act as black boxes and struggle with complex scenarios. To address this, recent works incorporate Vision-Language Models (VLMs) to provide explicit reasoning, enhancing both interpretability and driving robustness. These approaches typically rely on pre-generated annotations, which suffer from potentially flawed labels and require costly human labor. In this work, we propose a new framework that integrates structured reasoning and geometric precision through a teacher-student architecture. The teacher model introduces reflective reasoning, where the VLM generates logical explanations and then reflectively refines the reasoning under the supervision of ground-truth action. This enhances zero-shot generalization without intermediate labels. The student model distills the teacher's reasoning capabilities via supervised fine-tuning. We also design a separate waypoint decoder that interprets textual reasoning into continuous trajectories. Our proposed solution integrates two goals: providing explicit reasoning for interpretability and delivering robust and accurate driving performance. It leverages the synergy between these two objectives within a staged inference engine to enhance driving performance and explicitly uses the reasoning to guide driving prediction. Evaluated on Waymo benchmarks, our framework outperforms classical reasoning-based baselines in zero-shot reasoning, waypoint accuracy, and inference efficiency. Our experiments validate this design, demonstrating that the reasoning text makes a significant contribution to driving inference, resulting in around a 24% improvement in performance compared to an identical model that lacks reasoning. Our work advances reasoning-driven autonomous driving toward interpretable and deployable systems.

Multi-robot systems, particularly heterogeneous robot swarms, can improve the efficiency of complex task execution through parallel collaboration and complementary capabilities. However, conventional rule-based methods rely on predefined task models and specialized decision making programs, making it difficult to understand complex semantic instructions and coordinate heterogeneous robots. LLMs introduce strong language understanding and task reasoning capabilities, allowing multi-robot systems to interpret instructions, decompose tasks, and assign roles according to task semantics. VLMs further incorporate visual perception, enabling robots to reason about objects, regions, and spatial relationships in physical environments. Nevertheless, existing LLM/VLM based methods often depend on known maps, centralized and synchronized decision making, limiting their generalization to heterogeneous robots and unseen tasks. We therefore propose a framework that combines decentralized asynchronous reasoning, lightweight information sharing, capability aware collaboration, and a unified action interface, enabling general purpose VLMs to generate robot specific actions executed by learning free experts without task or robot specific training. Experiments across diverse scenarios and multiple VLMs show success rates above 70%, with completion time reduced by up to 55.8% relative to the geometric greedy baseline.

Vision-language models remain underused in colonoscopy despite the rich expert descriptions recorded in routine reports. These reports document lesion appearance, size and location but summarise entire procedures rather than caption individual frames, leaving clinical findings only weakly linked to the corresponding images. Here we develop EndoCLIP, a colonoscopy vision-language foundation model trained on 125,756 lesion-level image-text pairs progressively recovered from 280,476 routine colonoscopy records. Across lesion-level image-text retrieval, structured report generation and six multi-centre clinical classification tasks, EndoCLIP outperforms general-purpose and biomedical vision-language encoders in both zero-shot and linear-probe settings. On benign-versus-malignant classification, its linear probe approaches the performance of expert readers in a blinded study involving 12 endoscopists. These results suggest that recovering finding-to-frame correspondence can transform routine documentation into scalable supervision, enabling clinical targets to be specified in language rather than separately annotated for each task.

Detection and localization of AI-tampered images are critical for trustworthy AI, yet modern generative models have made such manipulations increasingly difficult to identify. While traditional binary classifiers can detect image tampering, they lack interpretability and generalization. Vision-Language Models (VLMs) offer a promising alternative due to their strong visual understanding and reasoning capabilities; however, existing approaches typically rely on supervised finetuning with curated explanations rather than exploiting their inherent reasoning capabilities. In this work, we investigate whether VLMs can be trained to reason about AI-generated image edits using reinforcement learning (RL) rather than explicit reasoning supervision. Motivated by the success in Group Relative Policy Optimization (GRPO), an RL technique that incentivizes the model to reason by generating thinking traces prior to giving the final answer, we propose a GRPO-based training framework that utilizes simple accuracy and format rewards. Given an input image, the model produces a structured reasoning trace and predicts whether the image has been tampered with. A lightweight segmentation model is then guided by the reasoning output to generate pixel-level localization masks. Experiments across multiple image manipulation datasets demonstrate that our approach achieves competitive detection and localization performance compared to state-of-the-art image forgery detectors, despite requiring substantially weaker supervision. We introduce effective intersection over union (eff-IoU), a unified metric to jointly evaluate detection and localization. These results suggest that reinforcement learning provides an effective and scalable mechanism for teaching VLMs to reason about AI-generated content.

Streaming video understanding requires models to continuously retain useful visual evidence before future questions are known. Existing approaches primarily manage the growing visual context according to token importance, temporal redundancy, or segment-level relevance, but rarely organize evidence around objects that persist and evolve over time. Thus, in this paper, we introduce ObjectStream, a training-free framework that treats latent objects as memory anchors for streaming video understanding. ObjectStream induces spatially coherent latent objects directly from frozen Video-LLM representations, links them across frames into persistent anchors, and maintains their histories under a bounded memory budget, without requiring external object detectors or segmentation models. Built on these anchors, ObjectStream preserves three complementary forms of evidence: persistent object histories, transient object changes, and recent visual context. This design enables existing Video Large Language Models (Video-LLMs) to reason over object identities, interactions, and state changes while leaving the underlying model unchanged. Extensive experiments on online streaming and offline long-video benchmarks demonstrate both effectiveness and efficiency. In online streaming evaluation, ObjectStream improves Qwen2.5-VL-7B by 10.0 points on OVO-Bench Real-Time Visual Perception, while reducing peak GPU mem-ory and TTFT by approximately 50%. On offline long-video benchmarks, it surpasses the full-token baseline while discarding 82.5% of visual tokens. These results highlight latent objects as a practical and effective organizing principle for compact streaming video memory.

Agentic vision-language models (VLMs), which interleave textual reasoning with explicit tool calls such as cropping and code-based image manipulation, have emerged as a compelling paradigm for reliable and interpretable multimodal reasoning. However, recent studies have revealed that such models often use tools unfaithfully. Many process images are irrelevant to the question (e.g., the tool crops the wrong region or misses the queried target), yet the call still receives full credit and the model still answers correctly. Such decorative or misaligned tool calls waste computation and reveal that the model leans on prior knowledge or the original image rather than the evidence it retrieves. This may stem from two limitations of prevailing methods: the tool reward fails to distinguish useful from useless calls, and tool feedback carries no signal of usefulness. To this end, we introduce FaithEyes, a multi-agent self-judging framework. Concretely, we use a VLM to judge whether each process image helps answer the question. The judgement is injected into the reasoning context as part of the tool observation to help subsequent reasoning, and meanwhile is used to scale the tool reward by the helpful-tool ratio to suppress reward hacking. To keep judgement available at evaluation and thus ensure train-test consistency, we further design a multi-agent framework where the model itself serves as a subagent to judge the tool calls from main agent, eliminating any dependence on an external model at inference. Training via a two-stage SFT + RL pipeline on adapted open-source data, FaithEyes attains competitive or superior accuracy across visual perception and reasoning benchmarks, while markedly improving tool faithfulness. The homepage is at https://github.com/Mosi-AI/FaithEyes.

As large vision-language models (LVLMs) are deployed globally, the combination of multilingual instructions and visual information makes malicious attacks more covert and sophisticated than ever before. However, existing methods isolate language and modality defenses, which, coupled with the scarcity of safety data and high fine-tuning costs, makes it difficult for models to defend against compound attacks. To address this severe challenge, we propose a neuron-level cross-dimensional safety alignment framework driven by modality- and language-shared safety neurons (MLS-Neurons). First, we identify monolingual and unimodal safety neurons by comparing responses to harmful and benign samples, quantifying functional saliency through activation strength and downstream impact. Then, by intersecting these unimodal neurons within each language, we extract modality-shared safety neurons (MS-Neurons) responsive to both visual and textual risks, bridging the safety representation gap between modalities. Furthermore, using English as a semantic anchor, we intersect MS-Neurons across languages to identify modality- and language-shared safety neurons (MLS-Neurons), serving as key defenses against compound attacks. Finally, we update only this minimal subset of shared neurons (~0.03% of parameters), transferring English-only safety supervision to multilingual and multimodal scenarios. Extensive experiments show that our method significantly outperforms state-of-the-art approaches across diverse multilingual and multimodal safety benchmarks while preserving general utility.

Large vision-language models (LVLMs) often hallucinate objects that are absent from an image. Despite recent progress, existing mitigation methods still lack reliable object-level grounding diagnostics and therefore tend to apply coarse-grained interventions, which can impair visual understanding, shorten responses, and reduce coverage of genuinely grounded objects. The key challenge is thus to detect, during generation, whether each emerging object mention is supported by reliable visual evidence, so that hallucination can be mitigated selectively. Yet output confidence reflects next-token plausibility rather than visual support, allowing language priors to make absent objects appear certain. We show that the missing diagnostic evidence is encoded in an Intrinsic Grounding Signature (IGS), a distributed signed attention pattern that remains informative for such confident hallucinations. Based on IGS, we propose Verifier-Guided Decoding (VGD), a decoding framework in which a lightweight verifier examines each emerging object mention, rolls back the KV cache when the mention is identified as high risk, suppresses the object and its synonyms, and regenerates the affected continuation. Because VGD intervenes only on object mentions identified as high risk, it reduces object hallucination while preserving the model's original visual understanding and grounded object coverage. Experiments on CHAIR and AMBER-G show that VGD achieves state-of-the-art object hallucination reduction: at @rec90, it cuts AMBER-G CHAIR by 43.6% while retaining 99.6% of grounded-object coverage, and reduces CHAIR-MSCOCO CHAIR$_i$/CHAIR$_s$ by 37.0%/30.4% without shortening captions.

Large vision-language models have improved at describing visual content, but accurate descriptions do not ensure interpretation when meaning depends on knowledge beyond the pixels. Memes expose this gap because they rely on cultural entities, background knowledge, and community conventions. Most meme benchmarks reduce interpretation to labels or holistic scores, obscuring where an explanation breaks down. We introduce MemeBench, a diagnostic benchmark of 1,253 Chinese and English memes with human-written references and quality-controlled VIKR annotations, centered on anime, comics, games, and adjacent online subcultures. Its VIKR schema decomposes explanations into Visual clues, Identity links, Knowledge units, and Reasoning mechanisms. Across 26 LVLMs, every model covers visible content more reliably than the knowledge needed to interpret it, and even the strongest retains a 22.6% Visual-Knowledge gap. To test whether this diagnosis can guide improvement, we introduce KAR, an entity-guided retrieval baseline built on CultureBase. Across four controlled models, KAR raises VIKR Success by 3.6-7.4% and, compared with generic retrieval, repairs more answers and breaks fewer. Yet both retrieval conditions improve Identity and Knowledge while reducing Visual coverage in every comparison. MemeBench reveals whether an interpretation succeeds, what is missing, and whether targeted evidence fills the diagnosed gap.

Large Vision Language Models have integrated reasoning capabilities, elevating cognitive performance to new levels. However, existing evaluations either focus solely on perception or rely on specific domains such as maths or coding. Evaluation for reasoning capabilities that align with an open-world environment is still required, especially one that considers perception and reasoning jointly. To bridge this gap, we propose to evaluate LVLMs by exploiting visual illusions as a diagnostic tool. Visual illusions are phenomena in which the human visual system misinterprets objective signals, resulting in an understanding that deviates from reality. We constructed IllusionReasoning, a benchmark of illusion images collected from the real world, incorporating diverse annotated question-answer pairs. Based on IllusionReasoning, we show that the reasoning capabilities of a wide range of LVLMs are not as advanced as claimed. Our work provides new insights into LVLMs and offers future direction for optimisation.

Large Vision-Language Models (VLMs) suffer from prohibitive inference overhead due to long sequences of visual tokens. However, existing visual token reduction methods mainly improve efficiency by pruning or compressing redundant tokens without examining whether the resulting representation remains semantically consistent with the original representation. Mapping the original N-token visual sequence to K tokens may discard, dilute, or misassign critical visual cues, triggering severe semantic drift that deviates the VLM's understanding. In this paper, we first introduce the principle of 'Calibrate Before Reason' to visual token reduction and propose CaRe, a training-free robust framework that calibrates compact visual representations before reasoning to preserve semantic fidelity in VLMs. CaRe consists of two mutually complementary modules: 1) Perturbation-Robust Calibration Anchoring, which identifies calibration anchors with stable model-side influence under multi-directional perturbations; 2) Confidence-Gated Token Calibration, which extracts reliable calibration signals from unselected tokens and injects them into anchors. Extensive evaluations across diverse VLM architectures and benchmarks verify that CaRe outperforms state-of-the-art token reduction baselines. While pruning 94.4% of visual tokens, our method retains 96.4% of the original full-token performance, delivering up to 2.30 times faster end-to-end inference speed relative to unpruned vanilla models.

Recent advances in large language models (LLMs) have led sign language translation (SLT), the task of converting sign-language videos into spoken-language text, to increasingly adopt LLMs as textual backbones. However, despite their strong language modeling capabilities, existing LLM-based SLT methods often undermine rather than exploit this language prior, producing disfluent translations, a failure we term language-prior degradation. Meanwhile, existing methods typically align videos and text at the sentence level, which does not ensure accurate lexical details and creates a lexical fidelity gap. To address both issues, we propose DualAnchor, a gloss-free LLM-based SLT training framework that couples two complementary anchors for linguistically fluent and visually faithful generation. Token-level Prior Anchoring (TPA) preserves the LLM's language prior by regularizing the multimodal decoder at each decoding step toward the next-token distribution of a frozen LLM conditioned on the same autoregressive prefix. Optimal Transport Alignment (OTA) improves lexical fidelity by formulating visual-textual matching as entropy-regularized partial optimal transport, with Sinkhorn optimization inducing a soft alignment between visual tokens and textual content tokens under a cosine cost. DualAnchor achieves strong overall performance on both PHOENIX-2014T and CSL-Daily. Targeted analyses attribute these gains to the complementary effects of the two anchors: TPA improves fluency, whereas OTA reduces fine-grained lexical errors.

Recent advances in Vision Language Models (VLMs) have created new opportunities for disaster response, where responders must interpret large volumes of sensor data under time pressure. Current VLM applications include social media monitoring for situational awareness, generation of draft action plans, and translation of technical alerts into public-facing messages. While these efforts can accelerate information flow, they remain largely limited to decision-support roles. Such approaches can increase operator burden because humans must still translate outputs into coordinated actions across teams and robotic assets. This study explores the viability of embedding VLMs as coordination agents within the human-UAV loop. The proposed architecture integrates natural language interaction, mission-level task coordination, software-in-the-loop implementation, and communication aligned with the Incident Command System (ICS). Rather than functioning solely as advisory tools, VLMs facilitate communication between human operators, mission control logic, and UAV task execution. The framework was developed using a Model-Based Systems Engineering (MBSE) approach, with use case and block definition diagrams representing system roles, internal structure, and component interactions. Three key elements, the VLM Coordinator Agent, UAV Mission Control, and Task Allocator, were implemented within an integrated simulation and control environment. A preliminary human-factors evaluation with seven participants showed reduced perceived workload across mental demand, effort, and frustration, along with high ratings for AI trust and communication clarity. By integrating MBSE, software-in-the-loop testing, and human-factors evaluation, this work advances scalable human-autonomy teaming for high-stakes disaster response, with broader implications for aerospace autonomy and civil safety.

Vision-language-action (VLA) models commonly adopt an LLM-centric $V \to L \to A$ pathway, where visual observations are projected into the representation space of a large language model before being decoded into robot actions. Although effective, this design incurs substantial computation and memory overhead at every policy invocation. In this work, we introduce TurboVLA, a new VLA paradigm that reformulates the conventional $V \to L \to A$ pathway as a direct $V + L \to A$ mapping. Instead of using a large language model as the central interface between perception and action, TurboVLA independently encodes visual observations and language instructions, directly exchanges information between them through lightweight bidirectional vision-language interaction, and predicts continuous action chunks with a compact decoder. This simple design constructs task-conditioned representations directly from visual and linguistic features, significantly reducing the computational and memory costs of VLA inference. On LIBERO, TurboVLA achieves 97.7% average success with only 0.2B parameters, 31.2 ms inference latency, and 0.9 GB inference VRAM on a consumer-grade RTX 4090, matching or outperforming substantially larger VLA policies. These results establish TurboVLA as a simple and effective alternative to the prevailing LLM-centric VLA paradigm, offering a new perspective on how vision, language, and action can be connected for efficient robotic manipulation. Code is available at https://github.com/H-EmbodVis/TurboVLA.

CT vision-language foundation models have demonstrated promising performance across downstream tasks, but are typically trained with whole-volume representations that dilute fine-grained anatomical signals. Fine-grained vision-language pre-training addresses this by aligning anatomy-level visual features with anatomy-specific text, but in doing so discards the global context that whole-volume models provide. Furthermore, existing fine-grained approaches train from scratch, making them computationally expensive. We introduce Anatomy Contextualized Adaptation (ACA), a lightweight framework that adapts frozen CT foundation model representations for anatomy-level vision-language alignment while enhancing global contextualization. ACA uses TotalSegmentator to decompose CT volumes into anatomy-level embeddings, which are refined via a transformer that captures cross-anatomy relationships, and aligned to both per-anatomy and scan-level text extracted from radiology reports. Evaluated on Merlin and CT-RATE, ACA consistently outperforms both the frozen foundation model baselines and existing fine-grained methods in zero-shot finding classification, while requiring less than one hour of training once embeddings are cached. The attention weights learned by ACA's inter-anatomy transformer additionally indicate plausible cross-anatomy context routing. Altogether, these results support ACA as a lightweight approach for adapting CT foundation models to anatomically grounded vision-language alignment while preserving and enhancing global anatomical context.

Gastrointestinal (GI) endoscopic image analysis has shifted from single-label classification toward visual question answering (VQA), where a model must answer free-form clinical questions about an image. While recent vision-language models (VLMs) achieve promising answer accuracy on this task, clinical adoption also requires the model's internal representations to reflect the visual evidence behind its answers. We propose a simple multi-task fine-tuning recipe that constructs auxiliary grounding and description tasks from an existing VQA dataset with minimal additional annotation: expert-annotated polyp masks are reused directly, while a GI-domain pretrained classifier with Grad-CAM localization provides weak supervision for finding categories that lack ground-truth masks. Three small VLM backbones are fine-tuned with low-rank adaptation under matched VQA-only and multi-task recipes on Kvasir-VQA-x1, and we show consistent accuracy gains together with improved implicit alignment between answer tokens and the relevant image region, evaluated on both in-distribution and out-of-distribution data.

Vision-language pre-training (VLP) serves as a cornerstone for medical multimodal representation learning. However, existing medical VLP frameworks are often constrained by the limited context windows and shallow representational capacities of lightweight text encoders when processing lengthy, terminology-dense clinical reports. While integrating medical large language models (LLMs) offers unprecedented clinical reasoning capabilities, it introduces three major bottlenecks: (i) the anisotropic representational collapse of generative LLMs under standard contrastive objectives, (ii) the prohibitive memory overhead of joint end-to-end training with large batch sizes, and (iii) the medical hallucinations induced by vanilla contrastive losses that ignore fine-grained anatomical laterality and negation modifiers. To address these challenges, we propose \textbf{SCALPEL}, a \textbf{S}emantic \textbf{C}ross-modal \textbf{A}lignment framework via \textbf{L}LM-\textbf{P}owered \textbf{E}ncoder \textbf{L}earning. First, Clinical Report Contrastive fine-tuning converts a generative LLM into an isotropic encoder via domain-specific clinical text adaptation. Second, an asymmetric alignment strategy leverages offline feature caching to enable efficient training. Critically, we formulate an Anatomy-Negation Aware Objective that explicitly penalizes mismatched image-text pairs involving laterality confusion or false negations. Extensive experiments across MIMIC-CXR, CheXpert, and IU X-Ray benchmarks demonstrate that SCALPEL achieves state-of-the-art performance in cross-modal retrieval, zero-shot disease classification and medical visual question answering.

Selecting a zero-shot out-of-distribution (OOD) detector for a new deployment is typically based on benchmark rankings, implicitly assuming that the highest-ranked detector will transfer across domains. We show that this assumption does not hold. Through a controlled portability audit across seventeen in-distribution datasets, three vision-language models, and seven representative zero-shot OOD detectors, we find that detector rankings reverse across deployments, every detector exceeds $80\%$ FPR95 on at least one domain, and the preferred detector depends on both the in-distribution data and the underlying VLM. We trace these reversals to complementary evidence channels in vision-language logits. Corpus-free detectors rely on different combinations of absolute match level and relative or spatial sharpness, while WordNet-based methods additionally depend on external semantic coverage. A simple proposition shows that level and sharpness cannot generally be recovered from one another, explaining why no single detector transfers reliably across deployments. Motivated by this diagnosis, we introduce the Complementary Evidence Guard (CEG), a detector-agnostic wrapper that preserves complementary evidence through a non-compensatory fusion of the base detector, level, and sharpness using only empirical in-distribution percentiles. Controls replacing these channels with entropy, logit variance, or random noise do not reproduce the gains. Without OOD samples, auxiliary corpora, or learned fusion, CEG reduces detector sensitivity and improves GL-MCM from $38.1$ to $28.8$ and MCM from $42.6$ to $30.5$ family-balanced FPR95.

Integrating 3D medical images with vision-language models (VLMs) holds substantial promise for computer-aided diagnosis. However, volumetric images generate prohibitively long visual-token sequences with considerable spatial and inter-slice redundancy. Existing token compression methods typically apply uniform reduction or rely on a single importance signal, increasing the risk of removing regions that are clinically relevant to the query or structurally distinctive. To address this limitation, we propose MedARC, a unified, training-free framework for Adaptive Redundancy Compression of visual tokens in 3D medical VLMs. MedARC estimates token importance by integrating three complementary cues: self-attention from the VLM vision encoder, which reflects the model's intrinsic visual focus; similarity between projected visual tokens and text embeddings, which identifies query-relevant regions; and deviations of local visual foundation model features from the volume-level feature center, which highlight structurally distinctive anatomy. The resulting importance distribution guides a saliency-aware merging strategy that preserves informative tokens while consolidating redundant ones rather than simply discarding them. Experiments on CT-RATE and MR-RATE show that MedARC reduces visual-token overhead and inference time while preserving or improving diagnostic performance. Its multi-cue scoring cost is outweighed by the savings from processing fewer tokens, with greater benefits expected for larger language models.

Reliable visual safety understanding in real-world scenarios demands more than just object recognition; it requires causal reasoning under epistemic uncertainty. While Large Vision-Language Models (LVLMs) demonstrate impressive semantic alignment on standard benchmarks, they often struggle to distinguish between superficial correlation and genuine forensic logic when grounded in the dynamic, partially observable nature of first-person experiences. Existing evaluations, dominated by third-person surveillance footage and binary classification metrics, fail to expose this cognitive gap. To address this, we introduce EgoSafe-Bench, a benchmark specifically designed to probe forensic reasoning in egocentric safety scenarios. It comprises 12,000 unique evaluation samples, generated by pairing each of the 3,000 video clips with a QA chain governed by our proposed Hierarchical Reasoning Evaluation (HRE) protocol. Unlike standard benchmarks, HRE mandates a rigorous reasoning trajectory from initial feature anchoring to blind-spot deduction and intent inference, thereby enforcing logical consistency and penalizing shortcut-based predictions. Extensive evaluations of state-of-the-art LVLMs (e.g., Qwen3-VL, Gemini, VideoLLaMA 3) reveal a significant perception-reasoning decoupling: models often achieve high descriptive scores but exhibit notable fragility in causal reasoning and logical closure. Our work provides both a challenging dataset and a systematic evaluation framework to foster the development of logically robust video understanding systems.

In this work, we study the use of Vision-Language Models (VLMs) for anomaly detection in an agent-driven game Quality Assurance (QA) pipeline focusing on geometry clipping. In this evaluation, a custom exploration agent navigates a game level to collect visual observations, while the automatic annotation pipeline provides frame-level clipping labels. This setup allows us to evaluate recent VLMs on a controlled anomaly detection task without manual annotation. We benchmark six recent VLMs (Gemini, GPT, Qwen, Gemma, Llama, and Ministral) under a zero-shot prompting setting and analyse their sensitivity to four prompt variants. Our results show that while the VLMs can capture visual cues associated with geometry clipping, they all produce substantial false positives on visually ambiguous frames such as near-contact geometry and partial occlusions. Gemini-3.1-Flash achieves the best overall accuracy and is the most robust to prompt variation, while open-source models exhibit large precision--recall swings depending on the prompt design. These findings suggest that current VLMs are best suited as high-recall candidate filters within multi-stage QA pipelines rather than as standalone bug detectors.

Fine-grained food visual--semantic understanding requires models to capture subtle distinctions across ingredients, cooking methods, doneness, color, texture, and plate composition. Although CLIP-style vision-language models provide a natural framework for this task, their effectiveness is limited when training relies on heterogeneous web-collected image--text pairs. Such data often exhibit a web-to-target domain gap and cross-modal misalignment, where images differ from the target distribution and captions are noisy, multilingual, or weakly grounded in visual content. We propose a data-centric multimodal alignment method for fine-grained food description and recognition. Our method first performs target-aware data selection to identify visually relevant training subsets, then applies VLM-based caption refinement to generate visually grounded, target-style descriptions. Using these curated image--caption pairs, we train complementary CLIP-style retrieval experts and further combine their decisions through a hierarchical VLM-assisted multi-expert decision-level fusion strategy that invokes the VLM only when experts disagree. Experiments show that our data refinement strategy significantly improves retrieval performance over naive web supervision, with VLM-based caption refinement alone yielding an average performance gain of approximately 19%. Our full method also achieves more than twice the retrieval score of pure VLM-based retrieval while remaining substantially more efficient.

Open-vocabulary segmentation labels arbitrary categories from a text query without per-class training, yet on remote sensing imagery it underperforms on categories it handles reliably elsewhere. We find that much of this gap traces to the text query rather than to the segmentation model. Because these models are not specialized for overhead imagery, the class name that serves as the query is often a weak address into the vision-language embedding space. We show that a better name repairs part of the gap, while the remaining failures call for an address that the tested natural-language rephrasings do not provide. We recover that address from a few examples through textual inversion on a frozen model, keeping inference text only. On a representative benchmark this raises the mean intersection over union on the affected categories from 3.9 to 39.4, and across eight remote sensing datasets it improves over few-shot methods that instead inject visual prompts at inference.

Unifying visual understanding and generation in one model holds immense promise, but remains challenging and expensive due to heavy compute and data demands and conflicts between the visual features needed for these two capabilities. To address these challenges, we present Argus-Unified, a compact, effective and unified multimodal model built with low demand on computation and data. Instead of aligning modalities from scratch, Argus-Unified effectively leverages pretrained vision-language models (VLMs) that provide strong multimodal priors. Specifically, we introduce hybrid visual tokens that preserve continuous tokens for understanding while learning discrete tokens for generation from a frozen unified vision encoder. Our training pipeline includes two stages: the first stage learns a quantizer and image decoder on top of the frozen vision encoder, the second stage trains the LLM initialized from a pretrained VLM for the unified multimodal modeling. Using by far the least amount of data (15.6M) and the lowest cost (~$2,000), we demonstrate that unified multimodal models can be trained economically while achieving strong performance in both understanding and generation. Notably, our model attains state-of-the-art multimodal understanding on GQA, POPE, and VQAv2, and competitive generation quality compared to models with dedicated vision encoders (e.g., Janus, Janus-Pro), all at ~10x lower cost and with ~5x less data. We envision Argus-Unified as a useful baseline that lowers the development barrier for unified models.

Vision--Language Models (VLMs) are increasingly deployed through a model supply chain in which pretrained checkpoints, architecture definitions, text encoders, and exported computation graphs are distributed by third parties and reused across downstream services. This reuse model creates a security-critical trust boundary: VLM deployments inherit not only learned parameters but also executable behavior encoded in shared model artifacts. In this paper, we show that a malicious provider can exploit this trust boundary by embedding architectural backdoors into VLM supply chains through representation steering. Our attack introduces dormant steering logic into the model architecture through a trigger-gated additive modification of an intermediate representation, without poisoning training data, controlling downstream fine-tuning, or modifying prompts at deployment time. When the trigger is absent, the modification reduces to zero and the model follows its normal computation, preserving clean utility. When the trigger is present, a steering direction shifts the internal representation toward an attacker-defined objective. We evaluate the attack across multiple VLM families and downstream tasks, including visual question answering, text-to-image generation, retrieval, and semantic response biasing. The results show that the proposed architectural steering backdoor compromises integrity, safety enforcement, and ranking fairness while preserving normal behavior on clean inputs. We further show that shared VLM artifacts can carry dormant steering logic against downstream services, and we propose an auditing defense that inspects the executable logic distributed with model artifacts rather than only their learned weights.

Zero-shot ObjectNav methods increasingly use vision-language priors, but direct object-object similarity in the latent space is often a weak proxy for spatial co-occurrence. We present an analytical, training-free semantic navigation pipeline that mediates object relationships through a compact room lexicon. Each object label is mapped to a CLIP-derived Room Probability Vector (RPV), and object-target co-occurrence is computed from RPV distribution overlap. These scores are projected onto a value map using geodesic flood-fill propagation (Fast Marching Method), with adaptive signal decay, and are used to rank frontiers by semantic score for navigation. Together, these components form an integrated, training-free, object-centric pipeline for open-vocabulary zero-shot navigation. Results show that our object-centric approach improves Success Rate (SR) and Success by weighted inverse Path Length (SPL) by a relative 3% and 1.3%, respectively, compared to image-holistic baselines on the HM3D dataset validation split, while preserving interpretability and open-vocabulary flexibility. Code is available at: uts-ri.github.io/RPV-SemNav.

Deep learning models have achieved state-of-the-art performance in several computer vision tasks. However, they experience severe performance degradation when applied to real-world scenarios due to unanticipated distribution shifts. Test-Time Adaptation (TTA) attempts to solve this problem by using unlabeled data from the target domain to dynamically adapt to the test distribution at inference time, without access to the source data. However, TTA remains a challenging problem when adapting to continuous, temporally correlated data, such as videos, and in scenarios where the target domain contains severe domain shifts. For this reason, few works in the literature explore TTA for videos under such extreme conditions. To overcome these limitations, we propose Test-time Adaptation via Dual Distillation (TADD), an online adaptation framework that relies on a lightweight projection adapter to bridge the domain gap. The adapter module is pre-trained on the source domain and then adapted to the target using our proposed complementary losses: (i) zero-shot distillation, which encourages alignment with the domain-agnostic features from a pre-trained vision-language model (VLM); and (ii) target distillation, which retains the source domain discriminative knowledge encoded in the pre-trained adapter. Built upon a frozen CLIP backbone, our method introduces this lightweight projection adapter as the sole updatable component during inference. We conducted extensive evaluations on three well-known video action recognition benchmarks: UCF-HMDB, Daily-DA, and Sports-DA. Our experiments in the closed-set scenario demonstrate that our method consistently outperforms state-of-the-art TTA baselines. Notably, our TTA approach improves upon previous methods by up to +3.81% on UCF-HMDB, +2.63% on Daily-DA, and +3.03% on Sports-DA.

Model merging aims to combine multiple domain-specialized experts trained from a shared foundation model into a single multi-task model. Existing approaches largely focus on improving the merging procedure itself and typically assume experts obtained through full-parameter fine-tuning. In this work, we revisit expert training for model merging. We first show that prompt-based adaptation provides a strong baseline: independently learned prompts can be exploited across tasks while keeping the backbone fixed, avoiding the interference introduced by weight merging. Building on this observation, we introduce Dual-Tuned Experts (DTEs), a two-stage training strategy that first learns prompts and then fine-tunes the vision encoder. This reduces the magnitude of task-specific parameter updates and produces experts with higher merge compatibility. Experiments across multiple CLIP architectures, full fine-tuning, and LoRA experts show that DTEs consistently improve merged performance of standard merging approaches and remain effective even when combining heterogeneous sets of experts.

Vision-language models (VLMs) deployed on consumer hardware must decide when to answer and when to defer, and that decision depends on having a confidence signal that tracks correctness. A practitioner with a fixed memory budget faces a choice between a small model at full precision, the same small model quantized, and a larger model quantized into the same footprint -- three configurations that push the confidence signal in opposing directions. We measure, on identical inputs, how model scale and 4-bit quantization affect two confidence signals in the Qwen2-VL family: the confidence a model states in natural language, and its own mean token probability over the answer it generates. Across 5,700 predictions spanning six realistic photographic degradations at three severities, we find that scale sharply improves the model's internal uncertainty signal (mean error-detection AUROC 0.80 to 0.98 from 2B to 7B) while its verbalized confidence stays weak and often at chance (mean 0.61 to 0.69): the gap between what the model knows and what it says widens rather than closes with size. We find that 4-bit quantization is nearly free for accuracy (-1.6 points) but expensive for the confidence signal (internal AUROC 0.95 to 0.80, and the verbalized-confidence parse rate collapses from 99% to 64%). For a fixed memory budget the recommendation is therefore to prefer a larger quantized model over a smaller full-precision one: 7B-4bit gives both the best accuracy and the best uncertainty signal (internal AUROC 0.98) of the three configurations that fit. We frame the results as selective-prediction operating points so they translate directly into a deployment recommendation, and we argue that error-detection AUROC, not calibration error, is the metric that exposes the difference between the two signals.

Vision-language models commonly project all tokens produced by a pretrained vision encoder into a large language model. However, final-layer features can discard text, local attributes, and spatial relationships, while high-resolution inputs substantially increase context length and inference latency. We introduce \method, a Multi-scale Adaptive Vision Encoder. \method uses position-dependent gates to fuse shallow, intermediate, and deep features from a vision Transformer, preserving global semantics while enhancing edges, text, and local structure. It then performs question-conditioned token routing according to question relevance, local information content, global semantics, and spatial coverage, with a token budget that adapts to image complexity. To mitigate compression loss, we further introduce full-to-compressed representation distillation and a spatial diversity regularizer. In an illustrative simulation under a unified 7B language-model framework, \method reduces the average number of SigLIP-SO400M visual tokens from 729 to 146 (approximately 80.0%) and improves the mean score on VQAv2, GQA, TextVQA, ScienceQA-IMG, and MMBench by 2.2 percentage points, while reducing single-image time to first token from 228,ms to 129,ms. We provide the full model design and evaluation protocol. All reported numbers currently serve only as placeholders for paper organization and experimental design; formal claims require real training runs, independent replications, and official benchmark evaluation.

Online platforms increasingly rely on automated age estimation systems to enforce minimum-age policies. Focusing on vision-based models designed for this task, concerns arise regarding their robustness to simple appearance changes that underage individuals may use to bypass such systems, such as drawing a mustache or applying lipstick. In this work, we present a systematic study of age verification robustness by simulating visual alterations that can be easily achieved by underage individuals. We evaluate seven models, including vision, vision-language, and multimodal large language models, across three datasets and four manipulation types. Interestingly, under drawn beard stubble, up to 61% of True Negatives are flipped into False Positives. Furthermore, we investigate how different demographics are affected by such manipulations, finding that Indians are more affected by beard stubble manipulations, while females are more affected than males across all manipulations. Finally, we explore how these biases can be mitigated using bias mitigation methodologies in lightweight linear probe settings.

Recent Vision-Language Models (VLMs) have achieved remarkable success in visual understanding, driven by the growing availability of high-quality image-text pairs. However, the performance of VLMs often degrades in the video domain due to the essential need for temporal understanding and the scarcity of large-scale annotated video data. In this work, we focus on marine video understanding, which brings further challenges: first, it requires substantial domain expertise; and video VLMs usually struggle with localizing and interpreting critical information from marine videos, as the informative events are typically sparse, unpredictable, and unevenly distributed. To address these challenges, we carefully curate the first event-centric marine video understanding dataset called MarineEVT, which features 20K multi-task, video-level visual question-answering pairs spanning multiple dimensions of marine understanding and analysis. Meanwhile, based on MarineEVT, we decompose marine video understanding as an Event-centric Visual Tool-integrated Reasoning process EVT-R1 for short, where we leverage powerful visual tools to drive the model to localize and interpret critical information aligned with visual questions and human intent. To demonstrate its effectiveness, we compare EVT-R1 against 11 SOTA VLMs in different settings. EVT-R1 outperforms the top open-source and top commercial models by 5.22 and 11.09, respectively. MarineEVT and EVT-R1 lay the foundation for ecological discovery and marine education, fostering the development of VLMs capable of interpreting marine dynamics, reasoning about ecological interactions, and supporting sustainable ocean video understanding and analysis.

Autonomous vehicles (AVs) depend on Global Navigation Satellite Systems (GNSS) for localization and navigation, making them vulnerable to spoofing attacks that can covertly redirect vehicles or induce unsafe maneuvers. In this paper, we develop the first Vision-Language Model (VLM)-based framework for GNSS spoofing detection for autonomous vehicles by fusing front-camera visual data with in-vehicle sensor readings (e.g., speed, acceleration, yaw rate) against GNSS-derived maneuvers. Our approach introduces a three-stage fine-tuning process that first grounds visual cues, and then calibrates sensor data within a shared semantic space to detect discrepancies between predicted and GNSS-derived maneuvers across three attack scenarios. We also generated an independent real-world dataset by driving an instrumented vehicle on public roads in Tuscaloosa, Alabama, equipped with time-synchronized GNSS, IMU, and camera logs to validate cross-regional generalization of our fine-tuned model on unseen data from training data. On this dataset, we then generated intelligent spoofing attacks, including trajectory mirroring with road-network snapping for wrong-turn attacks, position freezing for overshoot scenarios, and drift generation for stop attacks. On this validation dataset, the zero-shot VLMs baseline F1-score ranges from 23% to 32%, whereas our fine-tuned model achieves an F1-score ranging from 94% to 95%. Results show that our VLM-based approach correctly classified every wrong-turn and stop attacks, and attains 88%-93% accuracy for overshoot attacks. Furthermore, we introduce an adaptive inference policy that reduces VLM invocations to 14% (~86% computational reduction) and yields 65ms-73ms per 4s window. These results point to a practical, on-road layer of defense that complements signal-level integrity checks with the use of VLMs.

Modern vision-language models (VLMs) increasingly rely on dynamic or high-resolution visual encoding, producing thousands of visual tokens that substantially increase downstream language-model inference cost. Existing token-reduction methods assess token utility through token-wise importance, query relevance, coverage, pairwise diversity, or subset-level objectives. Our key insight is to view visual token reduction through selected-span complementarity: instead of scoring a token in isolation or through pairwise relations, we assess how much of its feature is orthogonal to the span of the already retained subset. Based on this perspective, we propose Greedy Orthogonal Token Selection (GOTS), a training-free and query-agnostic method. At each step, GOTS selects the token with the largest residual energy orthogonal to the current retained span. This rule exactly maximizes the one-step augmented Gram determinant among candidate additions, giving each greedy step a precise local geometric guarantee for subset expansion. Across five high-resolution VLM backbones from the Qwen-VL and InternVL families and eleven diverse benchmarks, GOTS achieves higher average performance retention than the strongest evaluated baselines, and a controlled OCRBench study shows that it reduces model-side time-to-first-token after accounting for selection overhead. Code is available at https://github.com/newLLing/GOTS.

A robot carrying a persistent, behavior-annotated map faces two planning questions, and its memory answers only one well. The \emph{spatial-navigation} question -- how to walk around a room -- we address first and report a negative: building on Vision--Language--Motion Maps (VLMM), a behavior-aware planner cost cuts a planning-time objective by $\sim$35% over 28 AI2-THOR scenes, but under closed-loop execution the real benefit nearly vanishes ($\sim$4%) and an on-demand vision--language model (VLM) does as well. The \emph{resource-allocation} question differs: under a limited perception budget, what should the robot re-observe now to keep its map fresh? Framing re-perception as this attention decision, we show a persistent map's memory (change-history, or even just recency of last sighting) yields the best schedule (held-out), matching an oracle, while the memoryless VLM prior is poor. Because the schedule reallocates budget toward what matters, memory's benefit concentrates on the important objects ($\sim$1.6$\times$ the mean), and a downstream fetch task confirms fewer wasted trips; the gain grows with per-instance heterogeneity exactly as a Cauchy--Schwarz bound predicts -- it equals $\mathrm{Var}(\sqrtλ)$, the variance of root-volatility. With a real CLIP prior on rendered objects the advantage is $+21$--$26\%$. The map's distinctive value appears when the task is \emph{language-conditioned}: told what to track, VLMM grounds the relevant objects (open-vocabulary) and tracks their change (memory), beating even a strong relevance-weighted recency baseline ($+2.5\%$) -- so its motion channel adds value beyond a last-seen timestamp -- and an on-demand VLM ($+8.9\%$); neither language nor dynamics alone suffices. The map earns its keep not by telling the robot how to walk around a room, but by telling it what to pay attention to.

The deployment of embodied agents in self-driving laboratories could accelerate scientific discovery, yet their reliability is constrained by the irreversible and safety-critical nature of chemical experiments. Progress is further hindered by scarce failure data and the lack of fine-grained evaluation protocols. To address these challenges, we introduce LabRobFail, a failure-centric framework for learning and evaluating robotic failure analysis in chemical laboratories. LabRobFail-Sim injects controllable failures at the control, physics, and semantic levels, enabling the construction of LabRobFail-Data, which contains over 20,000 trajectories across 70+ task scenarios, five failure categories, and 11 fine-grained failure types. LabRobFail-Bench evaluates six capabilities spanning task understanding, failure detection, temporal localization, severity assessment, failure classification, and actionable correction. We further develop LabRobFail-VLM, a domain-specialized vision-language model that generates structured failure diagnoses and recovery instructions. On seen environments, it achieves 90.83% failure-detection accuracy and 77.21% temporal-localization accuracy, substantially outperforming general-purpose VLMs. When integrated as a real-time supervisor, it improves downstream task success rates by 4-16 percentage points, demonstrating the value of fine-grained failure understanding for closed-loop recovery and reliable laboratory autonomy. Our code and data are available at https://github.com/Su-ISE-2001/SciRobo

Large Vision-Language Models (VLMs) now act as agents in interactive environments, where success requires coherent reasoning and decision-making across turns. Although end-to-end training in agentic environments can improve such multi-turn decision-making abilities, current methods mainly rely on either token-wise optimization over concatenated token trajectories or turn-wise optimization with uniform within-turn credit. In this work, we establish theoretical formulations for the two levels of optimization and derive a hybrid advantage that serves both objectives. Furthermore, with an appropriate choice of discount factor and learning target, we prove that a unified critic model can estimate values for both turn-wise and token-wise. As such, we propose HyGAE, an actor-critic framework that jointly optimizes token- and turn-level objectives with the hybrid advantage and unified critic. We conduct extensive evaluations of HyGAE across five multi-turn decision-making environments, where it achieves an average success rate of 91% and a significant improvement of 10% over other methods. Furthermore, we provide an in-depth analysis showing that the exact analytic form of the hybrid advantage and return is crucial for optimization. Project Page: https://wx-zhang.github.io/hygae-web/.

Multimodal misinformation verification is challenging because misleading signals may come from different parts of a post and require different forms of evidence. LVLMs are well suited to this task, but their verification performance often depends on the inference procedure applied to each instance. Existing methods improve this procedure through stronger prompting, retrieval, or deliberation, but rarely retain the verification patterns learned from previous examples. We propose Verification-Notebook Learning (VNL), a non-parametric framework that learns an external verification procedure for a frozen LVLM before inference. VNL builds a compact notebook of decision principles, evidence cues, and recurring pitfalls from prior verification experience. The notebook remains fixed during inference and guides the verification of new examples. Rather than updating model parameters or storing demonstrations, VNL records learned knowledge in an artifact that can be inspected directly. Experiments show that VNL consistently outperforms a range of competitive baselines. Further analyses show that the Verification Notebook improves fine-grained source attribution while remaining compact and interpretable, providing an effective way to accumulate verification knowledge without model training.

Vision-and-Language Navigation in Continuous Environments (VLN-CE) requires agents to maintain long-horizon visual history for trajectory consistency while executing actions with low latency. Existing video-based VLN approaches typically struggle to satisfy both demands simultaneously. To address these challenges, we propose MemVLN, a novel VLN framework that achieves state-of-the-art performance with real-time inference efficiency (14 FPS). MemVLN utilizes a visual encoder to process continuous observations and a Large Language Model (LLM) to interpret instructions and generate actions. Central to our approach is an Episodic Memory management that applies pyramidal resolutions. This mechanism concentrates computation on immediate percepts while retaining compressed long-term history. Complementing to this design, we introduce Procedural Memory for fast action with a compact vocabulary of atomic mid-level actions to bypass auto-regressive decoding latency. Experiments on VLN-CE show that MemVLN-4B surpasses the baseline Qwen3-VL-4B architecture by 5.8% SR in R2R and 9.7% SR in RxR, while achieving a 7$\times$ speedup in inference latency.

Large Vision-Language Models (LVLMs) remain bottlenecked by massive computational footprints, precluding their deployment on resource-constrained edge devices. While efforts to compress LVLMs focus heavily on vision token reduction or smaller language models, the vision encoder is largely overlooked, typically deployed as a monolithic, computationally heavy feature extractor. Moreover, there is no previous effort that designs a vision encoder for LVLMs directly optimized for on-device latency. In this paper, we present UltraViT, a vision encoder for LVLMs, explicitly designed and optimized for on-device performance. Specifically, by taking into account real on-device latencies, we systematically design a pyramidal architecture that strategically integrates and adapts heterogeneous spatial mixers at the macro-block level. Furthermore, to pre-train UltraViT, we propose a novel two-stage generative pre-training strategy: cultivating rich spatial features via dense distillation, followed by direct generative supervision from a capacity-mixed frozen LLM. Compared to standard contrastive and SSL, we show that our pre-training is much more effective for achieving high-level semantic grounding for UltraViT needed for the subsequent generative multimodal alignment of LVLM training. Extensive experiments demonstrate that our on-device latency-informed design combined with our tailored training strategy establishes a new state-of-the-art for efficient LVLM encoding, significantly outperforming existing encoder-centric baselines while operating on-device at nearly 1.7xthe speed.

Auxiliary signal pathways in VLMs are routinely fitted with learnable gates so the optimiser can decide how much of the signal to admit. We find that the optimiser almost always decides on zero: across five injection designs, every gated pathway becomes behaviourally closed, with accuracy invariant to ablating the pathway at inference even when the gate parameter would nominally pass 30-45% of the signal. We attribute this suppression phenomenon to two regimes, a dead-gradient regime formalised through the caption-invariance of image-derived signals, and a negative-utility regime in which the auxiliary signal actively hurts the loss. Rather than fight suppression, we exploit it: we regularise LoRA fine-tuning with geometric auxiliary losses from hyperbolic visual relational graphs (IoA-driven entailment cones and angular repulsion on the Lorentz manifold), coupled only through the forward pass at training time and dropped at inference. Disaggregating GQA by question type exposes a clean dissociation. Three configurations without geometric losses at inference lose 2.85-3.39pp on relational questions while gaining ~1pp on attribute questions; a fourth that trains with the losses but infers through a soft prompt loses 5.14pp on rel for only +0.23pp on attr, so training-time regularisation alone does not protect relational accuracy without a geometric inference pathway. Configurations that keep the geometric pathway at inference preserve vanilla-level relational accuracy and match the attribute gain. Out of distribution on VSR, the RMS-prefix recipe preserves the spatial signal; stripping the geometric losses (G2) collapses VSR by 4.6pp, isolating them as the OOD source. A secondary result: embedding-norm alignment is necessary for generation-safe prefix injection, and learnable gates should be replaced with fixed, non-optional injection at matched scales.

Post-training enables vision-language models (VLMs) to understand human instructions and perform various downstream tasks. Current post-training methods usually rely on human-annotated data, distillation from external models, reinforcement learning with human feedback, or verifiable answers. This limits their ability to improve without external supervision. To tackle this, we propose NOPD (Noisy Student On-Policy Self-Distillation), a simple yet effective self-distillation approach that improves VLMs without any external models or ground-truth answers. Our key insight is that prediction discrepancies between clean and corrupted inputs naturally induce a self-supervision signal. In NOPD, the model learns from corrupted inputs while using its own predictions under clean inputs as token-level supervision. We show the effectiveness of NOPD on five visual reasoning tasks; it can match and even outperform reinforcement learning approaches or distillation from external models. Notably, when trained with 2.1K samples from Geometry3K, NOPD improves Qwen2.5-VL-7B by 20 points on its validation set. It also shows generalization on out-of-distribution test sets and achieves 7.4 point gains on MathVista. Furthermore, we demonstrate that NOPD is a general approach to enhance VLMs, achieving improvements across three models on 12 benchmarks.

End-to-end OCR systems based on vision-language models have achieved strong performance in complex document OCR, but their efficiency is limited by the large number of visual tokens produced from document images. Many of these tokens correspond to blank margins or visually redundant regions, yet directly applying generic visual token compression methods may remove OCR-critical fine-grained details. In this paper, we propose LayoutLite, a lightweight plug-and-play module for efficient document OCR. Instead of relying on explicit document layout detection, LayoutLite performs implicit layout analysis at the token level between the vision encoder and the language decoder. It aggregates multi-layer visual representations from the vision encoder, and predicts an importance score for each visual token with a lightweight scoring network. Low-information tokens are then removed before entering the language decoder while preserving the original spatial positional information of retained tokens. To train LayoutLite without human annotations, we cast token selection as a reinforcement learning problem and optimize it with a group-relative policy optimization objective driven by OCR output consistency, together with an auxiliary layout supervision signal to stabilize training. Experiments on OmniDocBench demonstrate that LayoutLite can substantially reduce visual token length and inference cost with negligible degradation in recognition quality. We further evaluate LayoutLite on two OCR-specialized VLMs, FireRed-OCR and Logics-Parsing-V2. Under up to 50% token compression, LayoutLite preserves almost the same score on both models while reducing prefill latency, FLOPs, and KV cache memory by over 40%, with only a small additional inference overhead. These results show that token-level implicit layout analysis is an effective and practical approach for accelerating VLM-based OCR systems.

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

MEMEs are widely used on the internet and often carry strong elements of sarcasm or irony. Understanding their hidden meanings typically requires a joint interpretation of text and vision. Existing methods focus on the dual-stream vision-language model to extract the visual and text simultaneously, which lacks background information and prior knowledge about the comprehensive explanation of MEME. One feasible option is to adopt chain-of-thought (CoT). However, the simple CoT approach lacks multi-perspective thinking, which may compromise the reliability of the resulting answers. Moreover, it often relies on shallow feature fusion, lacking the fusion of local details and fine-grained visual-prompt text alignment. This limitation prevents a deeper understanding of the intricate connections between the visual and the text. Herein, an enhanced vision-language multi-CoT (EVL-MCoT) approach is proposed to address these limitations. By promoting multi-CoT, EVL-MCoT enhances consistency and reduces bias in the decision-making process. Additionally, we design a prototype-guided and context-guided decoding framework, which incorporates visual prototypes to guide the fusion process and enables the model to align textual and visual information more precisely. We achieve promising results on the HatefulMemes and MultiOff datasets. The source code has been publicly released and is available at https://github.com/BGWH123/EVL-MCoT.

Vision foundation models are increasingly reused as frozen backbones for downstream visual recognition, making parameter-efficient adaptation a central problem. Prompt-based adaptation, including Visual Prompt Tuning (VPT), provides a lightweight way to specialize these models, but its layer-wise behavior remains poorly understood: performance is sensitive to prompt depth, placement, and task distribution, and gains on standard in-domain benchmarks do not always translate into robust generalization. We argue that this limitation is not solely an optimization issue, but a layer-wise information allocation issue: existing prompt-based methods lack principled control over what prompt-conditioned representations should preserve, suppress, and propagate across depth. Inspired by the Information Bottleneck principle, we introduce Prompted Information Bottlenecks (PIB), a framework that regularizes layer-wise compression-sufficiency trade-offs and promotes a more coherent cross-layer information path. The key idea is that effective adaptation should be minimal yet sufficient, retaining task-relevant local evidence in earlier layers while progressively discarding nuisance factors and redundant details in deeper layers. Extensive experiments show that PIB achieves strong performance across 34 datasets, reaching 92.1% on FGVC, 93.01% on HTA, and 77.33% on VTAB-1k, while tuning only 0.35% parameters on average across the main settings. Beyond benchmark accuracy, PIB helps explain the non-monotonic behavior of prompt capacity scaling, reduces shortcut reliance, and improves robustness under distribution shift and fine-grained recognition settings. These results position PIB as both a practical method and an information-allocation perspective for adapting frozen vision foundation models. Our code is available at https://github.com/itsnotacie/MM-26-PIB

As semiconductor manufacturing advances toward sub-2nm nodes, local place-and-route (P&R) design-rule violation (DRV) fixing is increasingly limited by complex rule interactions, dense multi-layer routing geometries, and foundry-specific constraints. While Large Language Models (LLMs) have recently demonstrated strong capabilities in EDA scripting and documentation, their application to visual layout understanding remains largely unexplored: diagnosing DRC violations from layout imagery demands precise geometric reasoning and foundry-specific rule knowledge absent from general-purpose VLM training. We propose SCALE, a framework with a self-supervised layout-generation stage for local DRV fixing at advanced nodes. Multi-layer layout geometry is serialized into structured text, and a fine-tuned language model learns to reconstruct randomly masked polygons from surrounding BEOL context alone without violation labels. At inference, natural-language rule constraints and high-temperature sampling steer generation toward diverse, violation-prone layout variants validated by an industrial signoff DRC checker, producing DRC-annotated layout--violation pairs used to fine-tune a domain-adapted DRC-VLM. This VLM provides rule-aware geometric guidance for local DRV repair, boosting state-of-the-art agents' solve rates by +12--25% (up to 97%) on 100 real sub-2nm cases spanning enclosure, spacing, width, and color-spacing violations.

While Large Vision-Language Models (LVLMs) exhibit strong perceptual capabilities, they remain vulnerable in visual reasoning tasks. Existing benchmarks largely focus on symbolic mathematical or scientific problems and simple vision-centric tasks, offering limited assessment of complex visual reasoning and logical consistency, a critical requirement for reliable reasoning systems. We introduce ConVBench, a complex vision-centric reasoning benchmark in which each image is paired with two logically equivalent questions across six categories: action and state, complex counting, spatial reasoning, causal and intent understanding, commonsense reasoning, and temporal perception. To complement this benchmark, we define two evaluation metrics, logical consistency and robust accuracy, that jointly assess both the correctness and consistency of model responses. We further present ConVLM, which improves LVLM reasoning through Group Relative Policy Optimization (GRPO)-based reinforcement learning with a novel consistency reward. This method leverages automatically generated logically equivalent question-answer pairs and a dual-reward design combining accuracy- and consistency-based signals, encouraging agreement between paired responses. The framework functions effectively with or without strict answer supervision.

A vision-language AI assistant returns its answer as a stream of generated tokens. Therefore, a safety guard that watches that answer has to keep up with the stream and stop a harmful reply before a user reads it. Recent vision-language guardrails instead generate a chain of thought before they issue a verdict. They believe that step-by-step reasoning yields a safer guard. This design makes the guard heavy and slow, since the model must decode many tokens for harmfulness detection. We pose the question of whether a vision-language guard really needs to reason in order to screen a response. We answer with a guard that has no chain. ResponseGuard reads a harmful verdict from a single pooled representation of the request, the response, and the image in one forward pass. Across a standard multimodal guardrail benchmark, our 2B ResponseGuard outperforms a recent 3B reasoning-based vision-language guard on response harmfulness detection, without any reasoning and at about 150 times lower time cost. On request harmfulness the reasoning guard retains an overall lead, and the remaining gap on both tracks sits on the image-only cells. We observe that the gap may stem from the frozen vision encoders that both designs use rather than from the missing chain. We have also found the reasoning guard directs almost none of its verdict attention to the image. Based on a single-pass detection, ResponseGuard can screen an answer sentence by sentence as it streams and stop a harmful answer before it finishes. For guarding the response of a vision-language model, a calibrated single-pass label may provide a sufficient safety signal. We fully release all source code, trained models, and datasets at https://github.com/ndb796/ResponseGuard.

Electroencephalography (EEG) models used for epilepsy are often limited to specific datasets and tasks. This limited approach can make it challenging to apply these models across different datasets or in various situations. However, recent studies in foundation models and self-supervised learning suggest that an adaptable EEG backbone could support a range of EEG related tasks. In this study, we have developed a multimodal EEG foundation model that combines a raw signal encoder based on the Mamba architecture, a Vision Transformer (ViT)-style encoder for time-frequency data, and a lightweight encoder for text, all within a shared embedding space. The pretraining process relies on several innovative techniques, such as masked modeling, cross-view contrastive alignment, and temporal consistency losses. These methods are designed to create rich, seizure-relevant representations without requiring labeled data. To assess the efficacy and generalization of our pretrained model, we fine-tuned it on the canonical CHB-MIT seizure detection benchmark and additional seizure detection datasets, and conducted extensive experiments comparing different model variants. On the standard CHB-MIT split, our best single model achieved an AUROC of 0.874, and an ensemble variant reached 0.878 AUROC, representing state-of-the-art performance on this benchmark. In addition to standard train-test splits, we evaluated performance under a leave-one-subject-out (LOSO) protocol, which is rarely reported in prior EEG seizure modeling work and highlights the difficulty of patient-independent seizure detection, with a mean LOSO balanced accuracy of 0.558 across 19 subjects. Across datasets and evaluation settings, our multimodal foundation model enabled robust seizure detection and straightforward adaptation to new seizure detection scenarios, while also supporting interpretable seizure localization.

Adversarial attacks against large vision-language models (LVLMs) serve as an effective means of assessing their robustness in cross-modal semantic understanding. Existing studies mainly focus on corrupting visual inputs to induce predefined erroneous responses in general vision-language tasks, whereas corresponding investigations in remote sensing fields remain largely underexplored. Compared with natural image understanding, remote sensing image interpretation requires joint reasoning over local discriminative cues and global scene context. This poses additional challenges to achieving transferable semantic manipulation toward specified responses under black-box settings. To tackle these challenges, we propose GeoThreat, a transferable targeted adversarial attack method against LVLMs for remote sensing image interpretation. Specifically, GeoThreat modulates adversarial representations in accordance with the target content at both conceptual and perceptual levels. The class tokens from surrogate image encoders are employed as conceptual representations, while perceptual representations are distilled from patch tokens of the adversarial example through collaborative importance estimation. Beyond merely rolling out attention scores across layers, we incorporate adversarial-target similarity gradients to more faithfully characterize the relevance of local visual cues to the intended semantic manipulation. The perceptual representations are then dynamically aligned with target patch tokens in a cross-attentive manner, facilitating the adaptation of local cues toward designated semantic details. Finally, adversarial perturbations are iteratively updated via ensemble-based joint optimization of conceptual calibration and perceptual adaptation. Extensive experiments across diverse LVLMs demonstrate the superiority of GeoThreat in both transferability and controllability.

Improving video captioning quality typically demands retraining large vision-language models, an expensive and often impractical requirement. Existing training-free alternatives instead ground captions in detected objects to curb hallucination, but apply only a single, fixed correction pass without prioritizing which objects matter most, leaving semantically significant content omitted. We propose a prominence-aware, iterative post-hoc rectification framework that overcomes both limitations without modifying the underlying captioning model's parameters: a lightweight scoring mechanism ranks detected objects by spatial saliency, temporal persistence, and relational dynamics, and an iterative, prompt-driven refinement loop uses this ranking to progressively inject missing yet contextually relevant objects into the caption over multiple rounds. We validate the framework on MSVD and MSR-VTT using object-grounded automatic metrics, a 110-participant human study, and qualitative comparison against ChatGPT and Gemini; in human evaluation, the framework raises perceived completeness by up to 48% and reduces hallucination by up to 45% relative to a strong pretrained captioning baseline, all without retraining or reference captions. These results position prominence-guided iterative rectification as a lightweight, scalable, and model-agnostic route to more complete and trustworthy video captioning, with direct relevance to accessibility, retrieval, and other multimedia understanding applications.

Large vision-language models are becoming increasingly dominant in 3D medical image interpretation, but we rarely know <i>which</i> internal units encode clinical findings or <i>where</i> that information lives in the representation. We first study this on a 3D chest vision-language model (Pillar-0) by probing its frozen vision embeddings. We show that (i) each radiological finding is encoded by a <i>sparse</i> set of ~10 vision-encoder channels that match full-feature classification performance and far exceed a zero-shot text prompting; (ii) turning off the channels tied to one finding, that finding's score collapses while unrelated labels stay stable; and (iii) the same sparse probe <i>replicates</i> on an architecturally unrelated 3D abdominal VLM (Merlin) suggesting a general property of frozen medical encoders. Our training-free concept channel probe (CCP) method, paired with a corpus-derived report template, outperforms published CT-CHAT on clinical efficacy and NLG metrics (F1 0.549 vs. 0.184; BLEU 0.483 vs. 0.373) at 22x lower latency. Our results provide a clear, reproducible characterization of how frozen medical encoders represent findings, demonstrating direct applicability across models.

BoN improves model outputs by sampling several candidates and selecting one with a proxy score, but it assumes that complete candidates can be evaluated reliably. Many vision-language tasks instead provide only partial verification: a finding, span, value, region, or relation may be checkable even when no dependable whole-response verifier exists. Moreover, the same claim may recur across candidates with opposing stances, allowing one observation to support part of the pool and contradict another. We introduce Best-of-Evidence (BoE), an inference-time selection framework that keeps the BoN candidate pool fixed, represents reusable claims with a signed candidate--factor graph, and allocates a limited budget to evidence actions that can change the final choice. BoE formalizes selection under partial verification and provides a practical score-based controller, with the zero-budget case recovering the underlying BoN decision. Theoretically, we show that residual evidence capacity limits any evidence-driven improvement and that shared factor queries can achieve an O(log K) versus Θ(K) query separation in a factor-code model. Common-ledger experiments on four medical VQA settings show that BoE can improve fixed-pool selection and rescue some BoN failures when evidence is reliable, contrastive, and decision-relevant, while also revealing the channel-quality and candidate-generation limits that prevent universal gains.

Deploying navigation systems at scale requires a recipe that minimizes sensor assumptions, generalizes across robot embodiments, and trains efficiently. Yet, today's best systems depend on depth sensors, multi-camera rigs, or pre-built maps, limiting the hardware they support and increasing deployment cost. We introduce Robostral Navigate, an 8B vision-language model built around this scalability objective. The model consumes only a stream of monocular RGB images - the most ubiquitous sensor across robotic platforms and predicts waypoints by pointing to the next target location in the current camera view. Operating purely in image space, rather than robot-specific coordinates, makes the policy naturally robust to changes in camera intrinsics and scene scale, enabling deployment across wheeled, legged, and aerial robots without recalibration. We generate 2.4 million trajectories across 350k simulated scenes to reduce the reliance on real-world data collection and scale easily. We further introduce a prefix-caching training recipe that packs entire episodes into single training sequences, reducing training tokens by 22x and cutting training time from months to days. A tree-based attention mask prevents conditioning on previous ground-truth actions, encouraging visually grounded action prediction, and reinforcement learning is used to further improve exploration and recovery capabilities. On the Room-to-Room and Room-Across-Room in Continuous Environments (R2R-CE and RxR-CE) benchmarks, Robostral Navigate sets a new state of the art. On R2R-CE, it achieves a 77.4% success rate, surpassing the best monocular method by 10.5 points and the strongest depth- or multi-camera system by 5.3 points despite using only a single RGB camera. On RxR-CE, it reaches 75.1% success rate, outperforming all monocular baselines.

Small language models and coding agents increasingly generate web front-end code, yet their outputs are typically evaluated primarily for functional correctness. A generated interface may compile, render, and pass unit tests while still violating established interface quality principles, including accessibility barriers, deceptive design patterns, poor visual hierarchy, and excessive decision complexity. Existing auditing approaches face a trade-off between cost, coverage, and scalability: expert human review provides rich judgment but is slow and expensive; frontier vision-language models offer broader reasoning capabilities but remain costly to deploy at scale; and rule-based tools such as axe-core and Lighthouse are inexpensive but primarily capture mechanically checkable accessibility issues. We investigate whether a lightweight vision-language model can serve as an effective critic for generated interfaces. We unify 19 interface-quality principles from three complementary sources of HCI knowledge: WCAG 2.2 accessibility standards, deceptive design taxonomies, and established theories of perception, cognition, and interaction. To train this critic, we construct a verified dataset of approximately 10,000 generated web pages by synthetically injecting known violations into clean, LLM-generated Tailwind pages. Continued reinforcement learning on a 4B vision-language model improves micro-F1 from 36% to 84%, with 13 of 19 principles exceeding 80% F1. The resulting critic can audit generated interfaces, filter low-quality interface training data, and provide a reward signal for design-aware code generation. We release our data-generation recipe and injection/verification prompts to support reproducible evaluation and future work on scalable interface-quality assessment.

Understanding instrument-tissue interactions is essential for context-aware surgical AI and autonomous robotic surgery. Pretrained vision-language models (VLMs) and vision encoders offer an alternative to conventional interaction classifiers by transferring broad visual and semantic knowledge. However, adapting them to fine-grained surgical interactions remains challenging: (1) freezing the vision encoder depends entirely on pretrained representations that may retain noise and provide weak spatial localization, while (2) full fine-tuning can improve global semantic alignment without ensuring that the encoder learns meaningful features in the correct action region. We address these limitations by introducing LAViFiT, an end-to-end latent-action-guided framework for vision-language fine-tuning. An inverse dynamics model captures the visual changes induced by each action, while a forward world model drives the encoder to represent action-relevant regions. A patch-level SIG Regularizer further prevents local feature collapse without additional supervision, such as bounding boxes or pseudo-labels. Experiments across multiple encoders and datasets improve recognition and image-text alignment, while representation analyses show stronger grounding over the complete instrument-tissue interaction region and more spatially coherent features.

Vision-language models have shown strong potential for social robot navigation by leveraging rich semantic understanding of complex environments and human behaviors. However, large scale VLMs are difficult to deploy on resource-constrained robotic platforms, while lightweight VLMs often lack sufficient social reasoning capability. To address this problem, we propose SOPD-SocialNav, a selective on-policy distillation (SOPD) method that transfers social navigation knowledge from a large teacher VLM to a lightweight student VLM. SOPD introduces an entropy-based token selection mechanism that uses teacher uncertainty to identify socially informative decision tokens, while suppressing gradients from low-entropy tokens corresponding to trivial navigation states. A temperature-controlled Jensen-Shannon divergence objective is then used to align the student and teacher distributions on the selected tokens. Experiments on the SNEI and MUSON benchmarks demonstrate that SOPD consistently outperforms supervised fine-tuning, off-policy distillation, and standard on-policy distillation baselines in action prediction, perception consistency, and reasoning consistency. Real-world deployment on a Scout Mini robot further shows that the distilled model can generate more socially appropriate navigation behaviors in conversational and queuing scenarios. These results suggest that SOPD is an effective strategy for building lightweight yet socially aware VLM-based navigation systems.

Despite advances in Video Large Language Models (VLLMs) that have displayed promising outcomes in video understanding, the redundancy in the long-duration frames remains a hindrance to efficient reasoning. This paper introduces a training-free $\mathbf{P}$ersistence-Aware $\mathbf{C}$ompression and $\mathbf{A}$ggregation (PCA) method designed to preserve high-fidelity raw visual information before the encoding stage. PCA can be built on arbitrary VLLMs and consists of two modules: 1) A Dynamic Downsampling (DD) module that adaptively removes redundant frames by analyzing frame-wise similarity. 2) A Persistence-Aware Motion Enhancement (PAME) module that enriches each selected keyframe by aggregating the temporal context of its neighbors, ensuring that essential information is preserved even after aggressive frame reduction. Our approach substantially reduces the computation of long-context modeling, while enhancing the performance of the baseline model. Extensive experiments demonstrate that PCA consistently outperforms existing state-of-the-art approaches in both efficiency and accuracy, achieving a speedup of 1.8$\times$ to 2.5$\times$ compared to the baseline VLLM. The code is open-sourced at https://github.com/Heisenberg10110/PCA.

Visual information extraction (VIE) from visually rich documents remains challenging due to high layout variability and real-world impairments. Existing methods typically rely on sequential OCR pipelines or end-to-end models requiring extensive labeled data and layout-specific training, limiting their scalability.We propose a classification-guided large vision-language model (LVLM) framework for multi-type VIE that achieves high accuracy with minimal supervision. The approach decouples document-type classification from content extraction and employs in-context learning (ICL)-based dynamic prompt engineering to inject task-specific knowledge, enabling robust zero-shot inference across diverse layouts. From a theoretical perspective, the proposed method can be viewed as a form of conditional computation that reduces task uncertainty and improves information efficiency during prompt-based inference. Evaluated on a real-world bidding dataset with 16 certificate types, our zero-shot method (based on Qwen2.5-VL-7B) outperforms a strong supervised baseline by 18.35 percentage points in F1-score (86.43% vs. 68.08%) and 0.23 in normalized edit distance (0.90 vs. 0.67). Optional domain-specific fine-tuning further improves performance to 93.65% F1 and 0.93 NED, demonstrating superior robustness against seals, watermarks, and low contrast. The framework offers an efficient, scalable solution for complex document understanding in office automation. Code is available at https://github.com/FairmeHIT/Multi-VIE, and fine-tuned models at https://huggingface.co/fairme/Qwen2.5-VL-7B-SFT.