视觉大模型 / VLM

The rapid growth of ego-centric dashcam footage presents a major challenge for detecting safety-critical events such as collisions and near-collisions, scenarios that are brief, rare, and difficult for generic vision models to capture. While multimodal large language models (MLLMs) demonstrate strong general reasoning ability, they underperform in driving contexts due to domain and temporal misalignment. We introduce VLM-AutoDrive, a modular post-training framework for adapting pretrained Vision-Language Models (VLMs) to high-fidelity anomaly detection. The framework integrates metadata-derived captions, LLM-generated descriptions, visual question answering (VQA) pairs, and chain-of-thought (CoT) reasoning supervision to enable domain-aligned and interpretable learning. Off-the-shelf VLMs such as NVIDIA's Cosmos-Reason1 7B (CR1) exhibit near-zero Collision recall in zero-shot settings; fine-tuning with VLM-AutoDrive improves Collision F1 from 0.00 to 0.69 and overall accuracy from 35.35% to 77.27%. VLM-AutoDrive offers a scalable recipe for adapting general-purpose VLMs to safety-critical, temporally localized perception tasks. Evaluated on real-world Nexar dashcam videos, it achieves substantial gains in Collision and Near-Collision detection while producing interpretable reasoning traces, bridging the gap between perception, causality, and decision reasoning in autonomous driving.

Recently, vision-language models (VLMs) have made significant progress in 3D scene understanding, driving advances in applications such as embodied intelligence and robotic vision. However, existing approaches typically either rely directly on explicit 3D inputs (e.g., point clouds or RGB-D sequences), or introduce an additional 3D geometry encoder to derive 3D-aware visual tokens from 2D images. Such designs structurally decouple 3D geometric perception from the rich 2D semantics learned via vision-language pre-training, hindering the development of a unified 3D vision-language representation. In this work, we propose Occ-VLM, a novel framework for 3D scene understanding that operates purely on posed RGB images and employs a single 2D vision encoder. Specifically, Occ-VLM reconstructs 3D scene occupancy as an auxiliary geometric prior, which is utilized to spatially associate foreground 2D tokens with 3D space. These tokens are then decoded by a Large Language Model (LLM) for unified scene understanding. Extensive experiments demonstrate that Occ-VLM achieves both accurate geometric perception and robust vision-language reasoning: it attains state-of-the-art performance on multi-view occupancy prediction, while performing on par with 3D-input VLMs on 3D Visual Question Answering (VQA) and 3D dense captioning benchmarks.

Conventional vision-language models (VLMs) struggle to interpret scenes captured under adverse conditions (e.g., low light, high dynamic range, or fast motion) because standard RGB images degrade in such environments. Event cameras provide a complementary modality: they asynchronously record per-pixel brightness changes with high temporal resolution and wide dynamic range, preserving motion cues where frames fail. We propose RE-VLM, the first dual-stream vision-language model that jointly leverages RGB images and event streams for robust scene understanding across both normal and challenging conditions. RE-VLM employs parallel RGB and event encoders together with a progressive training strategy that aligns heterogeneous visual features with language. To address the scarcity of RGB-Event-Text supervision, we further propose a graph-driven pipeline that converts synchronized RGB-Event streams into verifiable scene graphs, from which we synthesize captions and question-answer (QA) pairs. To develop and evaluate RE-VLM, we construct two datasets: PEOD-Chat, targeting illumination-challenged scenes, and RGBE-Chat, covering diverse scenarios. On captioning and VQA benchmarks, RE-VLM consistently outperforms state-of-the-art RGB-only and event-only models with comparable parameter counts, with particularly large gains under challenging conditions. These results demonstrate the effectiveness of event-augmented VLMs in achieving robust vision-language understanding across a wide range of real-world environments.

While Multimodal Large Language Models (MLLMs) excel in general vision-language tasks, their application to remote sensing change understanding is hindered by a fundamental "temporal blindness". Existing architectures lack intrinsic mechanisms for multi-temporal contrastive reasoning and struggle with precise spatial grounding. To address this, we first introduce Delta-QA, a comprehensive benchmark comprising 180k visual question-answering samples. Delta-QA unifies pixel-level segmentation and visual question answering across bi- and tri-temporal scenarios, structuring change interpretation into four progressive cognitive dimensions. Methodologically, we propose Delta-LLaVA, a novel MLLM framework explicitly tailored for multi-temporal remote sensing interpretation. It overcomes the limitations of naive feature concatenation through three core innovations: a Change-Enhanced Attention module that systematically isolates and amplifies visual differences, a Change-SEG module utilizing Change Prior Embedding to extract differentiable difference features as input for the LLM, and Local Causal Attention to prevent cross-temporal contextual leakage. Extensive experiments demonstrate that Delta-LLaVA decisively outperforms leading generalist MLLMs and specialized segmentation models in complex change deduction and high-precision boundary localization, establishing a unified framework for earth observation intelligence.

Visual Question Answering (VQA) in robotic surgery, referred to as surgical VQA, requires high-level understanding of complex surgical scenes and the integration of visual perception with language reasoning, with the potential to support surgical training and intraoperative decision-making. Recent Vision-Language Models (VLMs) have shown promising performance through parameter-efficient fine-tuning; however, most existing approaches rely on coarse visual grounding, typically limited to bounding boxes, which fails to capture the fine-grained spatial structure of surgical objects. In this work, we propose a unified framework that jointly performs pixel-level segmentation and visual question answering within a single framework. Our approach integrates a VLM with a Segment Anything Model (SAM)-based decoder and represents scene elements as object tokens generated by the VLM. These object tokens guide answer prediction and are further projected to the SAM-based decoder to produce segmentation masks. By optimizing the object token embeddings through both segmentation and question answering objectives, the model learns spatially grounded representations that enhance visual reasoning while providing explicit pixel-level grounding. We evaluate the proposed method on the private RAMIE (Robot-Assisted Minimally Invasive Esophagectomy) dataset and the public EndoVis18 dataset, where it consistently outperforms baseline methods for surgical VQA. These results demonstrate that incorporating context-aware object tokens into vision-language models improves fine-grained surgical scene understanding.

We present a visual-context image-retrieval-augmented generation (ImageRAG)- assisted AI agent for automatic target recognition (ATR) of synthetic aperture radar (SAR) imagery. SAR is a remote sensing method used in defense and security applications to detect and monitor the positions of military vehicles, which may appear indistinguishable in images. Researchers have extensively studied SAR ATR to improve the differentiation and identification of vehicle types, characteristics, and measurements. Test examples can be compared with known vehicle target types to improve recognition tasks. New methods enhance the capabilities of neural networks, transformer attention, and multimodal large language models. An agentic AI method may be developed to utilize a defined set of tools, such as searching through a library of similar examples. Our proposed method, SAR Retrieval-Augmented Generation (SAR-RAG), combines a multimodal large language model (MLLM) with a vector database of semantic embeddings to support contextual search for image exemplars with known qualities. By recovering past image examples of known true target types, our SAR-RAG system can compare similar vehicle categories, thereby improving ATR prediction accuracy. We evaluate this through search and retrieval metrics, categorical classification accuracy, and numeric regression of vehicle dimensions. These metrics all show improvements when SAR-RAG is added to an MLLM baseline method as an attached ATR memory bank.

Manual annotation of high-quality visual question answering with grounding (VQA-G) datasets, which pair visual questions with evidential grounding, is crucial for advancing vision-language models (VLMs), but remains unscalable. Existing automated methods are often hindered by two key issues: (1) inconsistent data fidelity due to model hallucinations; (2) brittle verification mechanisms based on simple heuristics. To address these limitations, we introduce AutoVQA-G, a self-improving agentic framework for automated VQA-G annotation. AutoVQA-G employs an iterative refinement loop where a Consistency Evaluation module uses Chain-of-Thought (CoT) reasoning for fine-grained visual verification. Based on this feedback, a memory-augmented Prompt Optimization agent analyzes critiques from failed samples to progressively refine generation prompts. Our experiments show that AutoVQA-G generates VQA-G datasets with superior visual grounding accuracy compared to leading multimodal LLMs, offering a promising approach for creating high-fidelity data to facilitate more robust VLM training and evaluation. Code: https://github.com/rohnson1999/AutoVQA-G

Most production-level deployments for Visual Question Answering (VQA) tasks are still build as processing pipelines of independent steps including image pre-processing, object- and text detection, Optical Character Recognition (OCR) and (mostly supervised) object classification. However, the recent advances in vision Foundation Models [25] and Vision Language Models (VLMs) [23] raise the question if these custom trained, multi-step approaches can be replaced with pre-trained, single-step VLMs. This paper analyzes the performance and limits of various VLMs in the context of VQA and OCR [5, 9, 12] tasks in a production-level scenario. Using data from the Retail-786k [10] dataset, we investigate the capabilities of pre-trained VLMs to answer detailed questions about advertised products in images. Our study includes two commercial models, GPT-4V [16] and GPT-4o [17], as well as four open-source models: InternVL [5], LLaVA 1.5 [12], LLaVA-NeXT [13], and CogAgent [9]. Our initial results show, that there is in general no big performance gap between open-source and commercial models. However, we observe a strong task dependent variance in VLM performance: while most models are able to answer questions regarding the product brand and price with high accuracy, they completely fail at the same time to correctly identity the specific product name or discount. This indicates the problem of VLMs to solve fine-grained classification tasks as well to model the more abstract concept of discounts.

Video Question Answering (VQA) requires models to reason over spatial, temporal, and causal cues in videos. Recent vision language models (VLMs) achieve strong results but often rely on shallow correlations, leading to weak temporal grounding and limited interpretability. We study symbolic scene graphs (SGs) as intermediate grounding signals for VQA. SGs provide structured object-relation representations that complement VLMs holistic reasoning. We introduce SG-VLM, a modular framework that integrates frozen VLMs with scene graph grounding via prompting and visual localization. Across three benchmarks (NExT-QA, iVQA, ActivityNet-QA) and multiple VLMs (QwenVL, InternVL), SG-VLM improves causal and temporal reasoning and outperforms prior baselines, though gains over strong VLMs are limited. These findings highlight both the promise and current limitations of symbolic grounding, and offer guidance for future hybrid VLM-symbolic approaches in video understanding.

Multimodal large language models (MLLMs) frequently hallucinate by over-committing to spurious visual cues. Prior remedies-Visual and Instruction Contrastive Decoding (VCD, ICD)-mitigate this issue, yet the mechanism remains opaque. We first empirically show that their improvements systematically coincide with redistributions of cross-modal attention. Building on this insight, we propose Attention-Steerable Contrastive Decoding (ASCD), which directly steers the attention scores during decoding. ASCD combines (i) positive steering, which amplifies automatically mined text-centric heads-stable within a model and robust across domains-with (ii) negative steering, which dampens on-the-fly identified critical visual tokens. The method incurs negligible runtime and memory overhead and requires no additional training. Across five MLLM backbones and three decoding schemes, ASCD reduces hallucination on POPE, CHAIR, and MMHal-Bench by up to 38.2 percent while improving accuracy on standard VQA benchmarks, including MMMU, MM-VET, ScienceQA, TextVQA, and GQA. These results position attention steering as a simple, model-agnostic, and principled route to safer, more faithful multimodal generation.

Vision-Language-Action (VLA) models widely adopt pretrained Vision-Language Models (VLMs) as policy backbones, yet it remains unclear what kind of pretrained VLM representation is useful as a VLA initialization. In this paper, we study VLA initialization as a controlled representation-design problem along three axes: capability-level embodied VQA supervision, parameter-update strategy, and robot-data pretraining. Our experiments show that the original pretrained VLM representation is a key source of action performance. However, embodied VQA adaptation does not yield uniform gains: its benefit depends on downstream bottlenecks, and gains from different capability domains are not simply additive. For update strategy, LoRA provides a more reliable initialization than Full Finetune, indicating that overly reshaping the pretrained representation can weaken VLA initialization. Robot-data pretraining further improves VLA initialization, with the strongest variant obtained by staged LoRA-based training. Together, these findings suggest that effective VLM-to-VLA adaptation should inject action-relevant embodied and robot-trajectory signals while preserving the pretrained VLM representation that remains useful for action learning.

Recent advancements in Surgical Visual Question Answering (Surgical-VQA) and related region grounding have shown great promise for robotic and medical applications, addressing the critical need for automated methods in personalized surgical mentorship. However, existing models primarily provide simple structured answers and struggle with complex scenarios due to their limited capability in recognizing long-range dependencies and aligning multimodal information. In this paper, we introduce Surgical-LVLM, a novel personalized large vision-language model tailored for complex surgical scenarios. Leveraging the pre-trained large vision-language model and specialized Visual Perception LoRA (VP-LoRA) blocks, our model excels in understanding complex visual-language tasks within surgical contexts. In addressing the visual grounding task, we propose the Token-Interaction (TIT) module, which strengthens the interaction between the grounding module and the language responses of the Large Visual Language Model (LVLM) after projecting them into the latent space. We demonstrate the effectiveness of Surgical-LVLM on several benchmarks, including EndoVis-17-VQLA, EndoVis-18-VQLA, and a newly introduced EndoVis Conversations dataset, which sets new performance standards. Our work contributes to advancing the field of automated surgical mentorship by providing a context-aware solution.

Visual Question Answering (VQA), as the representative multimodal task, serves as a key benchmark for evaluating the reasoning capabilities of Multimodal Large Language Models (MLLMs). However, existing evaluations largely rely on static datasets and accuracy-based metrics, which fail to capture robustness, consistency, and generalization. Inspired by Metamorphic Testing (MT), we propose Metamorphic Robustness Assessment (MetaRA), a testing framework that employs Metamorphic Relations (MRs) to systematically probe vulnerabilities in MLLM-based VQA systems. MetaRA generates controlled variations of image-question inputs based on specific MRs and evaluates models across diverse conditions. Applying MetaRA to multiple MLLM-based VQA models across different tasks reveals nuanced failure patterns, including sensitivity to linguistic perturbations, over-reliance on superficial visual cues, and deeper weaknesses in multimodal reasoning. Experimental results demonstrate that MetaRA provides richer diagnostic insights than conventional accuracy metrics, exposing failure modes that remain hidden under standard benchmarks. Overall, this work highlights the need for systematic robustness evaluation in VQA and positions metamorphic assessment as a scalable, model-agnostic approach toward trustworthy multimodal AI.

Vision-Language Models (VLMs) have shown promise in various 2D visual tasks, yet their readiness for 3D clinical diagnosis remains unclear due to stringent demands for recognition precision, reasoning ability, and domain knowledge. To systematically evaluate these dimensions, we present DeepTumorVQA, a diagnostic visual question answering (VQA) benchmark targeting abdominal tumors in CT scans. It comprises 9,262 CT volumes (3.7M slices) from 17 public datasets, with 395K expert-level questions spanning four categories: Recognition, Measurement, Visual Reasoning, and Medical Reasoning. DeepTumorVQA introduces unique challenges, including small tumor detection and clinical reasoning across 3D anatomy. Benchmarking four advanced VLMs (RadFM, M3D, Merlin, CT-CHAT), we find current models perform adequately on measurement tasks but struggle with lesion recognition and reasoning, and are still not meeting clinical needs. Two key insights emerge: (1) large-scale multimodal pretraining plays a crucial role in DeepTumorVQA testing performance, making RadFM stand out among all VLMs. (2) Our dataset exposes critical differences in VLM components, where proper image preprocessing and design of vision modules significantly affect 3D perception. To facilitate medical multimodal research, we have released DeepTumorVQA as a rigorous benchmark: https://github.com/Schuture/DeepTumorVQA.

Visual Question Answering (VQA) models, which fall under the category of vision-language models, conventionally execute multiple downsampling processes on image inputs to strike a balance between computational efficiency and model performance. Although this approach aids in concentrating on salient features and diminishing computational burden, it incurs the loss of vital detailed information, a drawback that is particularly damaging in end-to-end autonomous driving scenarios. Downsampling can lead to an inadequate capture of distant or small objects such as pedestrians, road signs, or obstacles, all of which are crucial for safe navigation. This loss of features negatively impacts an autonomous driving system's capacity to accurately perceive the environment, potentially escalating the risk of accidents. To tackle this problem, we put forward the Dynamic Resolution Vision Language Model (DynRsl-VLM). DynRsl-VLM incorporates a dynamic resolution image input processing approach that captures all entity feature information within an image while ensuring that the image input remains computationally tractable for the Vision Transformer (ViT). Moreover, we devise a novel image-text alignment module to replace the Q-Former, enabling simple and efficient alignment with text when dealing with dynamic resolution image inputs. Our method enhances the environmental perception capabilities of autonomous driving systems without overstepping computational constraints.

Existing astronomy foundation models provide strong galaxy representations, but adapting them to new survey conditions and survey-specific morphology recognition tasks still requires substantial human supervision. We show that VLM-based VQA systems contain meaningful visual-semantic priors that can serve as weak supervision for downstream morphology classifiers and improve morphology classification under limited human-label budgets. We first introduce a survey-oriented VQA benchmark spanning two representative imaging regimes and evaluate state-of-the-art VLMs on galaxy morphology questions. The results show that these models capture useful morphology signals and informative uncertainty, but are not sufficiently reliable to replace human annotators. Motivated by this finding, we use a general-purpose VLM as a morphology teacher for Zoobot, an astronomy foundation model pretrained on large-scale Galaxy Zoo annotations. Across two survey domains and multiple annotation budgets, the VLM teacher consistently improves Zoobot's downstream morphology classification. These results demonstrate that a general-purpose VLM provides knowledge complementary to an astronomy foundation model and can teach it to better recognize galaxy morphology under limited human supervision. The resulting pipeline is designed for label-efficient adaptation to forthcoming large-scale surveys, including the Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST) and the Nancy Grace Roman Space Telescope. The benchmark and code are publicly available at https://github.com/fw-ic/VLM-morphology-teacher.

Vision Language Models have achieved near-human performance on single-document Visual Question Answering, yet their effectiveness degrades significantly when retrieving information from large collections of visually homogeneous documents. Existing multi-document benchmarks aggregate diverse document types, creating artificial separation in embedding space that does not reflect enterprise document repositories where thousands of records share identical visual templates. We identify this as embedding collapse and introduce Invoice Haystack, a benchmark with 1,500 anonymized invoice images paired with 200 discriminative question-answer pairs, specifically designed to stress-test retrieval under strong visual homogeneity. Invoice Haystack exhibits a mean pairwise cosine similarity of 0.73, compared to 0.38 (DocHaystack) and 0.31 (InfoHaystack) in existing benchmarks, posing a fundamentally more challenging retrieval problem. Addressing the identified challenge, we propose VL-RAG, a hybrid retrieval-augmented generation framework that jointly leverages text and visual embeddings to harness the complementary strengths of both modalities, followed by a VLM-based verification filter for precise document identification. VL-RAG achieves 60.0\% Recall@1 on Invoice Haystack-500, outperforming existing state-of-the-art method by up to an absolute 13.5 percentage points. It further improves retrieval considerably on DocHaystack-1000 (77.1\% vs.\ 75.2\%) and InfoHaystack-1000 (84.5\% vs.\ 80.0\%), establishing the proposed dual-stream fusion as a consistently superior retrieval strategy across both homogeneous and heterogeneous document collections.

Recent advances in 3D multimodal large language models (3D-MLLMs) have enabled unified solutions for 3D scene understanding tasks, including visual question answering, captioning, and referring segmentation. However, existing 3D-MLLMs remain largely object-centric, limiting their ability to model fine-grained part structures that are essential for embodied interaction with 3D environments. In this work, we present PAR3D, a unified part-aware 3D-MLLM framework that enables models to understand, reason about, and ground both objects and their parts in 3D scenes. To enable training and evaluation of part-aware 3D scene understanding, we introduce ScenePart, a synthetic 3D scene dataset with part-level annotations and language instructions. We further develop Part-Aware 3D Representation Learning to enrich 3D visual representations with fine-grained part-level semantics, and propose Hierarchical Segmentation Query Generation to ground part targets via hierarchical object-part queries. Extensive experiments show that our method substantially improves part-level question answering and referring segmentation, while also achieving strong performance across object-level vision-language tasks.

Vision-language models (VLMs) are prone to hallucination, which remains a major barrier to their safe deployment in clinical practice. To date, most hallucination detection methods have been evaluated on radiology benchmarks such as MIMIC-CXR and VQA-RAD, while gastrointestinal (GI) endoscopy remains largely underexplored. In this paper, we benchmark nine hallucination detection methods on the Gut-VLM dataset, a GI diagnostic Visual Question Answering (VQA) dataset with 4,392 test VQA pairs, across five VLMs (MedGemma-4B, MedGemma-27B, LLaVA-Med-7B, LLaVA-v1.6-7B, and Lingshu-32B). The methods span three categories: black-box methods (RadFlag, SelfCheckGPT-NLI), gray-box methods (AvgProb, AvgEnt, MaxProb, MaxEnt, Semantic Entropy, and VASE), and a white-box method (ReXTrust). Our results show that ReXTrust, a white-box method, achieves the highest AUC across all five models, outperforming the strongest alternative method on each VLM by a statistically significant margin (paired permutation test, p < 0.001 in all cases), reaching a peak AUC of 93.0 on MedGemma-4B. White-box hidden-state access provides a consistent advantage of 19.5 AUC points on average (range: 9.5--33.5), with ReXTrust maintaining strong performance even on LLaVA-v1.6-7B (AUC 79.9), where black-box methods and clustering-based gray-box methods collapse to near-chance performance. Among non-white-box methods, token-level gray-box statistics (MaxEnt, MaxProb) are the strongest alternatives, outperforming both clustering-based gray-box methods (Semantic Entropy, VASE) and black-box approaches on average. We further identify confident confabulation, a failure mode in which models hallucinate with high inter-sample consistency or high token-level probability, as a systemic failure for both consistency and uncertainty-based methods.

We present Jina-VLM, a 2.4B parameter vision-language model that achieves state-of-the-art multilingual visual question answering among open 2B-scale VLMs. The model couples a SigLIP2 vision encoder with a Qwen3 language backbone through an attention-pooling connector that enables token-efficient processing of arbitrary-resolution images. The model achieves leading results on standard VQA benchmarks and multilingual evaluations while preserving competitive text-only performance. Model weights and code are publicly released at https://huggingface.co/jinaai/jina-vlm .

Large vision language models (VLMs) have achieved impressive performance on medical visual question answering benchmarks, yet their reliance on visual information remains unclear. We investigate whether frontier VLMs demonstrate genuine visual grounding when answering Italian medical questions by testing four state-of-the-art models: Claude Sonnet 4.5, GPT-4o, GPT-5-mini, and Gemini 2.0 flash exp. Using 60 questions from the EuropeMedQA Italian dataset that explicitly require image interpretation, we substitute correct medical images with blank placeholders to test whether models truly integrate visual and textual information. Our results reveal striking variability in visual dependency: GPT-4o shows the strongest visual grounding with a 27.9pp accuracy drop (83.2% [74.6%, 91.7%] to 55.3% [44.1%, 66.6%]), while GPT-5-mini, Gemini, and Claude maintain high accuracy with modest drops of 8.5pp, 2.4pp, and 5.6pp respectively. Analysis of model-generated reasoning reveals confident explanations for fabricated visual interpretations across all models, suggesting varying degrees of reliance on textual shortcuts versus genuine visual analysis. These findings highlight critical differences in model robustness and the need for rigorous evaluation before clinical deployment.

Recent advancements in vision-language systems have improved the accuracy of Radiological Visual Question Answering (VQA) Models. However, some challenges remain across each stage of model development: limited expert-labeled images hinders data procurement at scale; the intricate and nuanced patterns of radiological images make modeling inherently difficult; and the lack of evaluation evaluation efforts makes it difficult to identify cases where the model might be ill-conditioned. In this study, we fine-tune a lightweight 3B parameter vision-language model for Radiological VQA, demonstrating that small models, when appropriately tuned with curated data, can achieve robust performance across both open- and closed-ended questions. We propose a cost-effective training pipeline from synthetic question-answer pair generation to multi-stage fine-tuning on specialised radiological domain-targeted datasets (e.g., ROCO v2.0, MedPix v2.0). Our results show that despite operating at a fraction of the scale of state-of-the-art models such as LLaVA-Med, our model achieves promising performance given its small parameter size and the limited scale of training data. We introduce a lightweight saliency-based diagnostic tool that enables domain experts to inspect VQA model performance and identify ill-conditioned failure modes through saliency analysis.

Vision-language models (VLMs) show promise for autonomous driving but often lack transparent reasoning capabilities that are critical for safety. We investigate whether explicitly modeling reasoning during fine-tuning enhances VLM performance on driving decision tasks. Using GPT-4o, we generate structured reasoning chains for driving scenarios from the DriveLM benchmark with category-specific prompting strategies. We compare reasoning-based fine-tuning, answer-only fine-tuning, and baseline instruction-tuned models across multiple small VLM families (Llama 3.2, Llava 1.5, and Qwen 2.5VL). Our results demonstrate that reasoning-based fine-tuning consistently outperforms alternatives, with Llama3.2-11B-reason achieving the highest performance. Models fine-tuned with reasoning show substantial improvements in accuracy and text generation quality, suggesting explicit reasoning enhances internal representations for driving decisions. These findings highlight the importance of transparent decision processes in safety-critical domains and offer a promising direction for developing more interpretable autonomous driving systems.

Many fields could benefit from the rapid development of the large language models (LLMs). The end-to-end autonomous driving (e2eAD) is one of the typically fields facing new opportunities as the LLMs have supported more and more modalities. Here, by utilizing vision-language model (VLM), we proposed an e2eAD method called SimpleLLM4AD. In our method, the e2eAD task are divided into four stages, which are perception, prediction, planning, and behavior. Each stage consists of several visual question answering (VQA) pairs and VQA pairs interconnect with each other constructing a graph called Graph VQA (GVQA). By reasoning each VQA pair in the GVQA through VLM stage by stage, our method could achieve e2e driving with language. In our method, vision transformers (ViT) models are employed to process nuScenes visual data, while VLM are utilized to interpret and reason about the information extracted from the visual inputs. In the perception stage, the system identifies and classifies objects from the driving environment. The prediction stage involves forecasting the potential movements of these objects. The planning stage utilizes the gathered information to develop a driving strategy, ensuring the safety and efficiency of the autonomous vehicle. Finally, the behavior stage translates the planned actions into executable commands for the vehicle. Our experiments demonstrate that SimpleLLM4AD achieves competitive performance in complex driving scenarios.

While Multimodal Large Language Models (MLLMs) offer strong perception and reasoning capabilities for image-text input, Visual Question Answering (VQA) focusing on small image details still remains a challenge. Although visual cropping techniques seem promising, recent approaches have several limitations: the need for task-specific fine-tuning, low efficiency due to uninformed exhaustive search, or incompatibility with efficient attention implementations. We address these shortcomings by proposing a training-free visual cropping method, dubbed FOCUS, that leverages MLLM-internal representations to guide the search for the most relevant image region. This is accomplished in four steps: first, we identify the target object(s) in the VQA prompt; second, we compute an object relevance map using the key-value (KV) cache; third, we propose and rank relevant image regions based on the map; and finally, we perform the fine-grained VQA task using the top-ranked region. As a result of this informed search strategy, FOCUS achieves strong performance across four fine-grained VQA datasets and three types of MLLMs. It outperforms three popular visual cropping methods in both accuracy and efficiency, and matches the best-performing baseline, ZoomEye, while requiring 3 - 6.5 x less compute.

Vision Language Models (VLMs) have demonstrated impressive performance in tasks requiring joint understanding of images and text, such as image captioning and Visual Question Answering (VQA), but our understanding of their internal processes remains limited. Recently, Sparse Autoencoders (SAEs) have emerged as a promising tool to support the interpretation of concepts encoded in VLMs. However, most SAE-based approaches focus only on textual or visual concepts separately, ignoring multimodal concepts. This limitation hinders a comprehensive understanding of VLMs, since concepts that integrate both modalities can be misclassified. Moreover, previous visual approaches often produce low-quality visual concept descriptions that are vague or incomplete, limiting their usefulness for understanding model reasoning. We propose a framework based on SAEs to extract and analyze visual, textual, and multimodal concepts from VLMs. For each neuron, we propose a candidate human-interpretable concept and compute the alignment between the concept and the dataset samples using cosine similarity scores. Experiments on a VQA dataset (LLaVA-NeXT) demonstrate that our framework improves visual concept quality by up to 45\% compared to existing SAE-based methods, while maintaining high textual concept quality and enabling systematic identification of multimodal concepts. This work contributes new insights into the conceptual space of VLMs, providing a structured approach to distinguish between visual, textual, and multimodal concepts. The code is available at https://github.com/PHDLanza/Multidata_SAE

Vision-Language Models (VLMs) leverage aligned visual encoders to transform images into visual tokens, allowing them to be processed similarly to text by the backbone large language model (LLM). This unified input paradigm enables VLMs to excel in vision-language tasks such as visual question answering (VQA). To improve fine-grained visual reasoning, recent advancements in vision-language modeling introduce image cropping techniques that feed all encoded sub-images into the model. However, this approach significantly increases the number of visual tokens, leading to inefficiency and potential distractions for the LLM. To address the generalization challenges of image representation in VLMs, we propose a lightweight, universal framework that seamlessly integrates with existing VLMs to enhance their ability to process finegrained details. Our method leverages textual semantics to identify key visual areas, improving VQA performance without requiring any retraining of the VLM. Additionally, it incorporates textual signals into the visual encoding process, enhancing both efficiency and effectiveness. The proposed method, SEMCLIP, strengthens the visual understanding of a 7B VLM, LLaVA-1.5 by 3.3% on average across 7 benchmarks, and particularly by 5.3% on the challenging detailed understanding benchmark V*.

Vision-language models (VLMs) are increasingly applied to medical visual question answering (Med-VQA), yet whether they can \emph{localize} the evidence behind their answers---a prerequisite for clinical auditability---is poorly characterized. We separately evaluate VQA reasoning and visual grounding for four recent frontier VLMs (GPT-5.1, GPT-5.5, Gemini-2.5-Pro, Gemini-3-Flash), two domain-specific medical VLMs (Lingshu, MedGemma), and a dedicated open-vocabulary detector (Grounding DINO) on VQA-RAD and SLAKE. Two findings challenge the intuition that ``add grounding to improve VQA.'' First, \textbf{no model localizes medical targets well}: every off-the-shelf system---frontier, medical-specialized, or dedicated detector---scores mean IoU 0.05--0.24 on our SLAKE grounding split, at or barely above a trivial center-box baseline (0.10), with Acc@0.5 below 20\%. Second, and counter to the common ``localize-then-answer'' paradigm, \textbf{cropping to a bounding box degrades VQA even when the box is a perfect oracle}: on the matched subset where oracle ground-truth boxes are applied, GT-grounding \emph{lowers} closed-ended accuracy for every model (by 0.9--18.0 points versus using the full image)---consistent with the crop discarding global context the model relies on. Because the oracle box removes localization error by construction, the problem is not that perception is a recoverable bottleneck, but that grounding-by-cropping is itself the wrong interface. Finally, we show constructively that the two channels need not conflict: supervised fine-tuning of Qwen-2.5-VL-7B on answers \emph{alone} silently destroys box-evidence emission (0/418 parseable boxes), whereas mixing in a small amount of grounding supervision restores localization to 0.36 IoU---above every zero-shot model---while preserving answer accuracy.

Vision-language models (VLMs) are increasingly used to detect whether AI-generated images contain visible artifacts, yet their ability to analyze such artifacts remains poorly understood. A correct image-level decision can still hide important failures: a model may correctly flag an artifact while relying on the wrong visual cue, selecting the wrong region, or describing a defect that the image does not support. To evaluate these behaviors directly, we introduce SalArt-VQA, a diagnostic benchmark for fine-grained SALient ARTifact understanding in AI-generated images. SalArt-VQA contains 950 images and 3,681 human-authored multiple-choice questions spanning artifact images, matched real reference images, and paired generated reference images. Four aligned question types evaluate presence detection, semantic localization, spatial grounding, and evidence-grounded defect identification, while the reference splits test calibration and abstention when the annotated defect is absent. Across 20 VLMs, SalArt-VQA reveals failures that image-level detection accuracy hides: the strongest model reaches 99.37% detection recall on artifact images but answers all four artifact-side questions correctly on only 53.26% of images. Comparing artifact images with artifact-free references reveals a sensitivity-calibration tradeoff: sensitive models often make unsupported artifact claims, while conservative models avoid false alarms largely by missing real artifacts. These results show that high artifact detection accuracy alone does not imply grounded artifact understanding. SalArt-VQA exposes these hidden failure modes and provides a fine-grained evaluation of whether VLM artifact claims are supported by local visual evidence.

Medical Vision-Language Models (VLMs) are typically evaluated on English radiology visual question answering benchmarks, leaving their robustness under non-English clinical language largely unexplored. We introduce IndoRad-VQA, an Indonesian adaptation of VQA-RAD, to assess whether medical VLMs retain radiology reasoning ability when questions are asked in Bahasa Indonesia. Radiology question-answer pairs are translated into Indonesian with self-evaluation-based quality control to preserve clinical meaning, terminology consistency, and answer equivalence. We evaluate general-purpose, Southeast Asian multilingual, and medical-specific VLMs under English and Indonesian prompting settings. Beyond accuracy, we quantify the language robustness gap between English and Indonesian inputs. We also conduct an error analysis to identify failure modes of question answering, such as yes/no flips, laterality errors, and output-language mismatches. Our findings show that strong performance on English medical VQA benchmarks does not necessarily translate to robust behavior in Indonesian clinical contexts. We observe a performance gap of 8 to 25 percent between the English and Indonesian settings, depending on the evaluation metric. These results highlight the need for more inclusive multilingual evaluation of medical multimodal foundation models. The dataset is available at https://huggingface.co/datasets/Lab-IS/IndoRad-VQA.