Multimodal Understanding

Follow this topic Ask about this topic

Momentum

10 papers in the last four weeks, up 233% on the four weeks before. 0.1% of all new papers.

Jul 13Week of Sep 28

Latest papers 413

Jul 8, 2026cs.AI

Infinity-Parser2 Technical Report

We present Infinity-Parser2, a large multimodal model that couples a controllable data-synthesis pipeline with multi-task reinforcement learning for end-to-end document parsing, addressing the persistent scarcity of faithfully annotated parsing corpora. Our contributions are threefold. First, we build a scalable synthesis engine, pairing a controllable rendering framework with an iterative refinement loop, and use it to construct and open-source Infinity-Doc2-5M: a 5-million-sample bilingual (Chinese/English) corpus spanning diverse document types, annotated with element bounding boxes, canonical content forms (Markdown, HTML, LaTeX, SMILES, structured charts), and full-page reading order. Second, we introduce a verifiable, multi-task reward system that enables Joint Reinforcement Learning across eight co-trained objectives (document parsing, layout analysis, table parsing, math formula parsing, chart parsing, chemical formula parsing, document VQA, and general multimodal understanding), unifying perception, structure, and reasoning in a single optimization signal. Third, we release two variants under a shared architecture: Infinity-Parser2-Flash, optimized for low-latency inference with a 3.68x throughput gain over Infinity-Parser-7B, and Infinity-Parser2-Pro, engineered for precision-critical settings. Infinity-Parser2-Pro reaches state-of-the-art 87.6% on olmOCR-Bench and 74.3% on ParseBench, surpassing DeepSeek-OCR-2, PaddleOCR-VL-1.5, and MinerU2.5, with strong generalization to charts, chemical formulas, and document VQA.

Zuming Huang, Jun Huang, Kexuan Ren +12

Synthetic Data Generation Multimodal Large Language Models Document Parsing Multi-Task RL Multimodal Foundation Models Multimodal Understanding

Jul 7, 2026cs.NE

Do You Remember? Toward Memory-Centric Multimodal AI

Human memory is reconstructive, not a faithful recording. Current multimodal LLMs (MLLMs) lack this capability: they process images through a frozen visual encoder, produce a one-shot text output, and discard internal representations. We present DoYouRemember, a three-stage architecture introducing reconstructive memory into MLLMs: (1) a VQ-VAE compresses images into discrete visual tokens, (2) a LoRA-fine-tuned LLM jointly attends to visual and text tokens, and (3) a Diffusion Decoder reconstructs images from the LLM's hidden states. On 1,000 3D facial skin texture maps and 99,000 unlabeled facial images, we find that LLM hidden states contain approximately zero recoverable visual information -- the same Decoder producing clear reconstructions from VQ-VAE tokens (pre-LLM) produces pure noise from LLM hidden states (post-LLM), demonstrating that the LLM understands images but does not remember them. Training a shared memory matrix M under backpropagation systematically fails due to gradient cancellation (O(1/sqrt(N)) attenuation). We identify three root causes and show that local EMA updating resolves all three: each image updates only its top-8 slots out of 64, preserving inter-slot diversity. The resulting M (229K parameters, 16x compressed) approaches the VQ upper bound on unseen test images. Scaling to 1,024 slots surpasses it (LPIPS 0.056 vs. 0.071), as M's continuous representation avoids VQ quantization error. We unify these findings under an information-theoretic framework: memory is lossy compression, recall is decompression, and hallucination is an inherent property of lossy decompression rather than a defect.

Xuguang Yu, Weigang Zheng, Minyue Yu

Memory-Augmented VLMs Multimodal Memory Memory-Augmented Language Models Neural Network Memorization Image Reconstruction Multimodal Large Language Models+3

Jul 7, 2026cs.CV

Vision as Unified Multimodal Generation

We formulate computer vision as unified multimodal generation, where heterogeneous visual tasks are expressed in the native text and image generation spaces of a unified multimodal model, without task-specific architectures. Under this formulation, SenseNova-Vision uses natural-language instructions and optional visual prompts to specify tasks, target regions or views, and decoding conventions, and generates responses as text for symbolic outputs, images for dense spatial predictions, or mixed text-and-image outputs for compositional tasks. To support large-scale training, we convert diverse computer vision annotations into instruction-response examples compatible with these generation spaces, resulting in the SenseNova-Vision Corpus, a computer-vision instruction-response corpus spanning text, image, and mixed targets. Starting from an off-the-shelf pretrained unified multimodal model, SenseNova-Vision is trained primarily on this corpus, with auxiliary multimodal data used as a capability-preserving mixture, and requires no task-specific prediction heads or architectural modifications. The resulting model covers a broad range of vision tasks, including detection, OCR, keypoint estimation, segmentation, depth estimation, surface normal prediction, point maps, and camera pose estimation, while supporting language-defined variants that combine category, color, region, and other visual cues. Experiments show that a single unified model can match leading task-specialized systems across structured visual understanding, dense geometric prediction, segmentation, and multi-view visual geometry. These results suggest unified multimodal generation as a scalable route for integrating computer vision capabilities into general-purpose foundation models. The model and corpus are publicly available.

Xiaoyang Han, Jianhua Li, Kewang Deng +14

SenseTime Research · Peking University · Shanghai Jiao Tong University +3

Unified Multimodal Models Multimodal Generation Multimodal Foundation Models Multimodal Understanding

Jul 7, 2026cs.IR

CMDR: Contextual Multimodal Document Retrieval

Multimodal document retrieval aims to retrieve relevant pages while preserving both textual and visual content from the original document. However, existing benchmarks primarily evaluate simple lexical or semantic matching, and most methods encode pages independently. Consequently, they overlook the contextual information in the document required to resolve queries that aggregate information across multiple pages. In this paper, we introduce CMDR and CMDR-Bench, a new multimodal document retrieval task and benchmark that require modeling document context. To address this challenge, we propose CMDR-Embed, a contextual multimodal embedding framework that explicitly incorporates document context by jointly encoding multiple pages and deriving page-level embeddings from a shared contextual representation. Furthermore, we introduce CMCL, a contextual multimodal contrastive learning objective that effectively trains CMDR-Embed by balancing contextual modeling with page-level discriminability. Experiments demonstrate that CMDR-Embed significantly outperforms non-contextual embeddings, highlighting the importance of context-aware multimodal embeddings for advancing document retrieval.

Ryota Tanaka, Taku Hasegawa, Kyosuke Nishida

Human Informatics Labs., NTT, Inc., Tokyo, Japan

Multimodal IR Multimodal Embedding Visual Document Retrieval Multimodal Contrastive Learning Multimodal Understanding

Jul 7, 2026cs.CL

Audio Sentiment Analysis via Distillation and Cross-Modal Integration of Generated Multilingual Transcripts

Automatically recognizing the sentiment, positive or negative, from speech is a challenging task, requiring both the analysis of vocal inflections and the interpretation of uttered words. Recent solutions rely on audio foundation models to solve the task, but it remains unclear if such models can take all aspects into account. To this end, we propose a multimodal solution that integrates audio and text information via cross-modal transformers, where text transcripts are automatically generated via an automatic speech recognition (ASR) tool. Moreover, we create multiple text modalities by automatically translating the transcripts into multiple languages via machine translation tools. Audio and multilingual text features are combined via a cascaded architecture comprising cross-modal transformer blocks that integrate modalities one by one. We further distill knowledge from the multimodal model, called teacher, into a unimodal (audio only) model, called student. We conduct experiments on a large-scale dataset, demonstrating that the automatically generated textual information can bring significant performance boosts in multimodal sentiment polarity classification. Our ablation study confirms that both automatic transcripts and automatic translations are helpful. Moreover, we show that the audio-only model can be enhanced via distillation, boosting performance without any computational overhead during inference. To reproduce the reported results, we publicly release our code at https://github.com/andreidurdun/cross-modal-audio-sentiment.

Andrei-George Durdun, Victor Constantinescu, Radu Tudor Ionescu

Department of Computer Science, University of Bucharest, 15G Iuliu Maniu, Bucharest 061075, Romania · Department of Data Science, PPC Romania, 30 Mircea Voda, Bucharest 030667, Romania

Cross-Modal Learning Sentiment Analysis Multimodal Sentiment Analysis Multimodal Understanding

Jul 6, 2026cs.CV

Light-Omni: Reflex over Reasoning in Agentic Video Understanding with Long-Term Memory

Agentic video understanding equips models with long-term memory to autonomously process and respond to continuous, long-horizon multimodal streams. However, advanced video agents often rely on ``detective-style'' iterative reasoning for action control (e.g.,

\mathtt{search}

) and evidence aggregation, incurring prohibitive costs and latency. We argue that such heavy reasoning primarily compensates for the lack of global context and semantic misalignment in retrieval. This paper introduces Light-Omni, a multimodal agent framework for reflexive and lightweight video understanding. It achieves this through dual contextual states that instantly build the required context in a single forward pass. First, we maintain a global state, a finite-sized multimodal script continuously consolidated from episodic memory, serving as the global context for Light-Omni. Through hierarchical merging, it preserves recent details while summarizing past events. Second, conditioned on this global context, we generate a parametric latent state that directly drives autonomous actions and produces retrieval embeddings, with minimal latency. Benefiting from this coupled design, Light-Omni achieves semantically aligned retrieval and reflexive responses while avoiding iterative reasoning. Extensive experiments validate the effectiveness of Light-Omni across multiple video benchmarks. Notably, it outperforms M3-Agent with an average 2.4% accuracy gain, a 12.1

\times

speedup, and a 2.6

\times

improvement in GPU memory efficiency. Furthermore, it serves as a memory system to enhance both the performance and efficiency of existing MLLMs. Project page: https://clare-nie.github.io/Light-Omni.

Chang Nie, Jiaju Wei, Junlan Feng +2

Nanjing University

Multimodal Agents Memory-Augmented Agents Video Retrieval Temporal Video Understanding Multimodal Understanding

Jul 6, 2026cs.CL

Unified Audio Intelligence Without Regressing on Text Intelligence

Audio intelligence involves understanding, reasoning about, and generating both audio and speech. In this work, we introduce Nemotron-Labs-Audex-30B-A3B (Audex), a unified audio-text LLM built on Nemotron-Cascade-2-30B-A3B, a strong text-only MoE LLM. Audex adopts a simple unified design with a single Transformer decoder: audio inputs are encoded and projected into the text embedding space, while text tokens and quantized audio output tokens are treated uniformly during generation. This architecture enables strong audio-text fusion, seamless multimodal generation, and compatibility with standard LLM training and inference infrastructure. For training, we meticulously curate audio-text datasets comprising 157.4B audio tokens and 320.5B text tokens. We apply multi-stage supervised training on these datasets, followed by text-only Cascade RL and multi-domain on-policy distillation. Audex delivers state-of-the-art audio understanding, speech recognition and translation, text-to-speech, audio generation, and speech-to-speech generation, while preserving very compelling reasoning, alignment, knowledge, long-context, and agentic capabilities of its text-only LLM backbone with marginal or no regression. We release the model checkpoints to facilitate open research.

Zhifeng Kong, Sang-gil Lee, Jaehyeon Kim +17

NVIDIA

TTS Synthesis Unified Multimodal Models Audio Understanding Audio-Language Models Automatic Speech Recognition Multimodal Foundation Models+1

Jul 6, 2026cs.CV

Enhancing Large Multimodal Models in Key Information Extraction via Scene-Aware Document Synthesis

Key Information Extraction (KIE) converts visually rich documents into structured data, but practical deployment remains challenging: strong performance often relies on costly on-server Large Multimodal Models (LMMs), while compact locally deployable models lack sufficient KIE supervision. We present SAYRE, a scene-aware document synthesis framework for generating scalable KIE training data without hand-crafted template design. Given a few exemplar documents, SAYRE captures category-specific content patterns and layout conventions to synthesize document-schema-annotation triples. It further introduces error-driven generation, which expands real-world failure cases into hard training examples while preserving their structural patterns. Experiments on constrained- and open-category KIE show that SAYRE consistently improves Qwen3-VL backbones and achieves the strongest overall performance among on-device LMMs. Data scaling experiments show an overall upward trend as more synthesized data is introduced, especially for smaller models and open-category extraction. Error analysis further shows that synthesized training reduces field-level errors by improving schema-aware extraction over dense tables, business identifiers, and contract clauses. These results establish scene-aware synthesis as an effective data-centric approach for improving practical multimodal KIE.

Zhipeng Xu, Zulong Chen, Qing Liu +6

Alibaba Group · Taobao and Tmall Group, Alibaba · Qwen Team, Alibaba Group +1

Synthetic Data Generation Document Information Extraction Multimodal Large Language Models LLM Information Extraction Multimodal Understanding

Jul 6, 2026cs.CV

Hierarchical Evidence-Driven Reasoning for Long Document Understanding

Retrieval-Augmented Generation (RAG) streamlines long-document understanding by leveraging retrieval mechanisms to restrict input images to a highly curated subset. However, existing multimodal RAG pipelines primarily face two critical challenges: first, standard semantic similarity retrievers frequently fetch topically overlapping yet answer-void distractor pages that mislead downstream generation; second, rigid single-pass pipelines heavily depend on initial retrieval success, where any omission of core evidence inevitably causes cascading errors. To address these challenges, we introduce HIEVI-RAG, a hierarchical, evidence-driven multimodal RAG framework for closed-domain document understanding. HIEVI-RAG systematically factorizes complex queries into a cooperative four-stage pipeline: (1) hierarchical question decomposition to break multi-hop root queries into atomic child questions; (2) coarse visual page retrieval leveraging a multimodal retriever to fetch candidate pages based on semantic similarity; (3) fine-grained page verification via EVIAGENT, a specialized multi-page verifier trained with GRPO to execute cross-page reasoning over multi-image blocks; and (4) memory-guided iterative generation that leverages accumulated sub-question context to execute multi-round, dynamic reasoning over the prioritized sequence. Extensive evaluations across four benchmarks demonstrate the robust efficacy and synergy of our framework, which significantly outperforms existing open-source baselines and exceeds the strongest reported baseline by an average of 8.05% in accuracy.

Junyu Xiong, Yonghui Wang, Rongjian Gu +4

University of Science and Technology of China · iFLYTEK Research

Multi-Hop QA Document Understanding Multimodal RAG Question Decomposition Question Answering Visual Document Retrieval+1

Jul 5, 2026cs.CV

Transferability Between Understanding and Generation in Unified Multimodal Models

Unified Multimodal Models (UMMs) integrate image understanding and generation within a single architecture, yet how the two tasks interact remains understudied. We investigate

\boldsymbol{\mathsf{transferability}}

in UMMs: whether training a capability on one task improves the same capability on the other without explicit supervision. Through controlled experiments, we empirically find that transferability depends on architecture-models with fully shared transformer backbone and a unified visual encoder exhibit consistent cross-task transfer, while loosely coupled designs show little or none. Leveraging this transferability, we propose a practical training strategy. The most straightforward way to improve a target generative capability (e.g., counting) is to fine-tune generation directly, but this can degrade visual quality due to distribution shift. Instead, we train the corresponding understanding task and let it transfer into generation, which improves capability-specific generative performance while minimizing distribution shift. We validate this across three capabilities-counting, spatial relation, and text recognition/generation-showing that cross-task transferability can be systematically exploited in UMMs.

Jiwon Kang, Heeji Yoon, Jaewoo Jung +5

KAIST AI, South Korea · Blynx, South Korea · Trillionlabs, South Korea

Unified Multimodal Models Multimodal Generation Transfer Learning Multimodal Understanding

Jul 4, 2026cs.CV

Q-TriM: Question-Guided Tri-Modal Attention for Audio-Visual Question Answering

Audio-Visual Question Answering (AVQA) extends classical VQA by requiring joint reasoning over video and synchronized audio. However, many AVQA systems rely on deeply stacked layers of self- and cross attention across text, video, and audio. Such sequential stacking may incur loss of information such as subtle inter-modal cues over the layers, causing errors to accumulate across sequential attention layers during the fusion. We introduce Q-TriM which performs multi-modal fusion in a shallow and parallel manner instead of a deep and sequential manner. For Q-TriM, we propose a novel framework for attention operation incorporating video and audio conditioned on text. As a result, we obtain not only standard cross attention outputs but also Tri-Modal Attention representations in which Query, Key, and Value come from distinct modalities. These attention representations are combined in parallel at a single stage, thus avoiding the multi-modal fusion with deep stacks in order to mitigate error accumulation and depth-induced issues. Q-TriM achieves state-of-the-art performance on three AVQA benchmarks, including substantial gains on MUSIC-AVQA-R, which demonstrates its robustness and out-of-distribution generalization. Code is available at https://github.com/Sunghun95/Q-TriM

SungHun Kim, SeungJun Baek

Dept. of Computer Science and Engineering Korea University Seoul, South Korea

Multimodal QA Cross-Modal Attention Audio-Visual QA Multimodal Fusion Multimodal Understanding

Jul 4, 2026cs.AI

Bridging Interleaved Multi-Modal Reasoning as a Unified Decision Process

Unified multi-modal models (UMMs) have shown promising interleaved text-image reasoning capabilities, yet effectively optimizing such multi-turn generation via reinforcement learning (RL) remains an open challenge. Existing approaches apply RL exclusively to text steps, relegating image generation to supervised surrogates, preventing policy gradients from propagating through the full interleaved trajectory across heterogeneous modalities. This leaves the potential of RL for UMMs largely untapped. In the paper, we introduce \textbf{BRAID} (\textbf{B}ridging inte\textbf{R}le\textbf{A}ved mult\textbf{I}-modal reasoning as a unified \textbf{D}ecision process), a simple framework that casts multi-turn text-image-text reasoning as a unified Markov decision process (MDP), enabling joint optimization of textual and visual generation via a single, principled RL objective. BRAID computes a shared trajectory-level advantage and propagates it coherently into both text tokens and image denoising paths, each optimized through its modality-native policy gradient mechanism. To further address long-horizon credit assignment, BRAID employs a vision-language model (VLM) judge that scores each intermediate image on its reasoning utility, supplying dense turn-level feedback to sharpen learning at critical visual branches. Experiments on spatial reasoning and visual perception benchmarks show that BRAID consistently outperforms various baselines, confirming that a unified MDP formulation with vision-thinking guidance is essential for effective multi-modal reasoning.

Zican Hu, Xuyang Hu, Yiming Liu +10

Nanjing University · Shanghai AI Laboratory · Tsinghua University +5

Unified Multimodal Models Multimodal Reasoning Multimodal Optimization Multimodal Generation Multimodal Foundation Models RL for VLMs+1

Jul 4, 2026cs.CV

Attending to Multimodal Generation One Token at a Time

Multimodal large language models (MLLMs) generate responses autoregressively, integrating visual and linguistic information in an evolving context. Prior work on interpretability has focused on individual layers and circuits (where), leaving the token-level dynamics of multimodal computation during generation (when) underexplored. We address this gap and study attention shifts as per semantic role; tracking model attention to image, text, instruction, and previously generated tokens, One Token at a Time (OTaT). We introduce multimodal tasks that require explicit switching between visual and textual context within a single response. Across two mainstream model families and four open-weight MLLMs of varying sizes, we establish consistent patterns: attention to image peaks at tokens requiring image-derived information, instruction tokens are revisited during task transitions, and attention to previously generated tokens increases as the generation progresses. Causal attention blocking interventions validate the functional role of these trends. We profile model behavior under disrupted attention and observe responses falling back to language priors, or exhibiting cross-modal leakage, denial, or recovery. Finally, informed of the attention dynamics through our novel analysis, we propose a simple test-time intervention to boost attention to the relevant modality at the right time, significantly improving multimodal task performance.

Varun Gupta, Vineet Gandhi, Makarand Tapaswi

CVIT, IIIT Hyderabad

Multimodal Large Language Models Multimodal Generation Multimodal Foundation Models Multimodal Model Interpretability Multimodal Understanding

Jul 3, 2026cs.IR

Modality Relevance is not Modality Utility: Post-hoc Selective Modality Escalation for Cost-Aware Multimodal RAG

Multimodal retrieval-augmented generation (RAG) grounds a generator in evidence drawn from heterogeneous modalities -- text, tables, and images. The dominant deployment choice is binary and made before the model has tried to answer: either run a cheap text(+table) pipeline, or pay for an expensive vision-language model (VLM) over every image. Recent adaptive systems improve on this by selecting the modality or fidelity pre-retrieval, from a question-conditioned predictor of which modality will be needed. We show that this is the wrong decision point. Through an oracle headroom analysis on MultiModalQA, we find that the relevance of a modality to a question is a weak predictor of whether that modality is actually needed to answer correctly: a large fraction of questions whose gold support includes an image are nonetheless answerable from text and tables alone, and a pre-retrieval router that escalates on apparent visual relevance over-escalates substantially relative to an oracle. We propose \textbf{post-hoc selective modality escalation}: answer cheaply from text and tables, run a verifier on the (query, draft answer, evidence) tuple that localizes which modality is missing, and pay for VLM evidence only there. A calibrated value-of-escalation router then decides whether the expected accuracy gain justifies the visual cost. On MultiModalQA, our router recovers the accuracy of an always-on VLM pipeline while issuing far fewer visual calls, and closes most of the gap to the oracle escalation rate. The result extends a routing-signal hierarchy established for retrieval depth and reasoning hops to a third axis -- modality -- under a single cost-aware selective-escalation view.

Xue Li, Yiming Gai

Hangzhou International Innovation Institute, Beihang University Hangzhou, China

Multimodal RAG Efficient Multimodal Inference Multimodal QA Multimodal Understanding

Jul 3, 2026cs.LG

OmniFocus: Query-Guided Modality-Balanced Token Compression for Omni-Modal Large Language Models

Omni modal large language models (OmniLLMs) have attracted wide attention for their ability to jointly process audio and video, but they generate large token sequences under audio-visual inputs, leading to substantial inference cost. Existing audio-visual token compression methods often rely on unimodal guidance, overlooking the temporal locality of query-relevant evidence in audio-visual inputs and implicitly assuming that the two modalities share a temporally aligned information density distribution. We propose \textbf{OmniFocus}, a training-free query-guided token compression method for OmniLLMs that performs independent importance estimation for video and audio, enabling a modality-symmetric compression design that preserves modality-specific salient evidence while maintaining audio-visual alignment, thereby mitigating the modality bias issue that can arise from unimodal-guided compression. Experiments on the Qwen2.5-Omni model family across four audio-visual benchmarks show that OmniFocus maintains strong compressed performance at low token retention ratios and outperforms existing baselines on several major benchmark scores at 25% token retention. On DailyOmni with Qwen2.5-Omni-7B at 25% token retention, OmniFocus maintains 59.40 accuracy while delivering up to 1.38

\times

prefill speedup relative to the full-token baseline, highlighting a favorable practical accuracy-efficiency trade-off.

Shijie Cao, Qingyu Zhang, Boxi Yu +6

School of Advanced Interdisciplinary Sciences, University of Chinese Academy of Sciences · Chinese Information Processing Laboratory, Institute of Software, Chinese Academy of Sciences · University of Limerick +1

Efficient Multimodal Inference Multimodal Large Language Models Multimodal Token Compression Multimodal Foundation Models Multimodal Understanding

Jul 2, 2026cs.CL

Gemma 4 Technical Report

We introduce Gemma 4, a new generation of open-weight, natively multimodal language models in the Gemma model family. Designed to advance compute efficiency and reasoning, the Gemma 4 model suite features dense and Mixture-of-Experts architectures, ranging from 2.3B to 31B parameters. Alongside improved vision and audio encoders for all model sizes, we propose a unified, encoder-free architecture for our 12B model, which ingests raw audio and image patches. Furthermore, we integrate a thinking mode, enabling Gemma models to generate reasoning traces prior to responding. We improve inference speed, memory, and compute efficiency, as well as long-context abilities through critical design choices. Gemma 4 establishes a leap in performance across STEM, multimodal, and long-context benchmarks, and rivals larger, frontier open models in human-rated tasks.

Gemma Team, Sherif El Abd, Vaibhav Aggarwal +298

Efficient Multimodal Inference Unified Multimodal Models Long-Context Language Model Inference Multimodal Large Language Models Multimodal Reasoning Open-Weight Language Models+2

Jul 2, 2026cs.CV

Show Me Examples: Inferring Visual Concepts from Image Sets

Vision-language models (VLMs) can follow complex textual instructions, yet they struggle to reason from purely visual context. In particular, current models fail to infer shared concepts from sets of example images and apply them to new inputs. We introduce Visual Concept Inference from Sets (VICIS), a task that evaluates this capability. Given a small context set of images sharing a concept and a query image, the model must generate new images that preserve the context-defined concept while remaining consistent with the query. We show that state-of-the-art VLMs perform poorly on this task, often ignoring the visual context or defaulting to biased generations. To address this gap, we propose a training framework and architecture that learn to infer visual concepts from image sets and extract concept-specific embeddings from queries. Experiments on synthetic data and large-scale ImageNet/WordNet data show that our model generates more accurate and diverse outputs and generalizes to unseen concepts and modalities such as sketches.

Nick Stracke, Kolja Bauer, Stefan Andreas Baumann +3

Munich Center for Machine Learning · Apple

Vision-Language Models Conditional Image Generation In-Context Learning VLM Reasoning Multimodal Understanding

Jul 2, 2026cs.AI

Hidden Forgetting in Continual Multimodal Learning: When Accuracy Survives but Grounding Fails

Multimodal large language models must continually adapt to evolving tasks and domains, yet standard continual learning metrics mainly measure whether old answers remain correct, leaving the stability of multimodal grounding largely unexamined. We study this overlooked failure mode and ask whether a continually adapted MLLM can preserve not only what it answers, but also how it uses visual, textual, OCR, chart, and document evidence. We identify \emph{hidden evidence-use forgetting}, where answer accuracy is retained while the model silently shifts toward different or less grounded evidence channels, and propose \textsc{RCL}, a replay-free reliance-constrained continual learning framework. \textsc{RCL} freezes the previous checkpoint as a behavioral reference, estimates teacher and student evidence-reliance profiles through counterfactual channel interventions, and jointly optimizes task learning, prediction preservation, and reliance preservation without adding inference-time cost. Across CoIN, COAST, MCITlib, and an evidence-sensitive multimodal stream, \textsc{RCL} consistently improves final performance and reduces forgetting over replay-free, PEFT, routing, and memory-assisted baselines, while substantially lowering modality reliance drift, dominant evidence flips, and hidden forgetting rates. These results suggest that robust continual multimodal learning requires preserving the evidence path behind correct answers, not merely the answers themselves.

Qianyu Chen, Canran Xiao, Runxuan Tang

1Nanyang Technological University · 2Shenzhen Campus of Sun Yat-sen University

Multimodal Grounding Multimodal Foundation Models Continual Learning for VLMs Multimodal Understanding

Jul 1, 2026cs.CV

Towards High-Resolution Visual Perception via Hierarchical Entity Exploration

High-resolution (HR) image perception remains a key challenge in multimodal large language models (MLLMs), as fine-grained details are often lost when the image is processed as a whole. Existing methods either require training to teach models where to look or heuristically divide the image into fixed regions, both of which struggle to generalize in complex HR scenes. In this work, we propose Hierarchical Entity Exploration (HEE), a training-free and model-agnostic framework that transforms static image understanding into dynamic, query-guided entity exploration. HEE first evaluates each region using a dual scoring mechanism to determine whether it already contains sufficient evidence to answer the question. If not, it applies object detection within the most promising region to extract fine-grained entities, clusters them into coherent subregions, and organizes them into a multi-level semantic hierarchy for deeper exploration. When deeper regions still fail to yield confident answers, a confidence-guided backtracking mechanism revisits alternative paths to ensure adaptive perception. Extensive results show that HEE outperforms training-free methods like ZoomEye and RAP in both accuracy and efficiency on two complex HR benchmarks (Visual Probe and HR-Bench), across different MLLMs such as Qwen2.5-VL and LLaVA-OneVision. Moreover, HEE demonstrates generalization on the MME-RealWorld benchmark.

Ziyu Ma, Shidong Yang, Yuxiang Ji +5

AMAP, Alibaba Group, China · Monash University, Australia

Large Vision-Language Models Fine-Grained Visual Perception Multimodal Understanding VLM Inference

Jul 1, 2026cs.MA

M2Note: Continual Evolution of Vision Language Models via Mistake Notebook Learning

Vision Language Models (VLMs) have demonstrated remarkable capabilities in multimodal reasoning tasks, yet they still suffer from recurring failures, such as skipping key visual checks, misapplying domain rules, and hallucinating unsupported concepts. Most existing solutions rely on supervised fine-tuning (SFT) and reinforcement learning (RL), which are expensive to iterate and can be brittle under distribution shift. To this end, we propose Multimodal Mistake Notebook Learning (M2Note), a training-free continual evolution framework that externalizes learning into an editable memory. M2Note transforms failed trajectories into compact subject-guidance notes: the subject summarizes the underlying domain and concept, while the guidance provides actionable verification steps that can be reused in future inference. At test time, M2Note retrieves relevant notes via multimodal retrieval-augmented generation (RAG) and appends them to the model context, steering reasoning away from previously observed pitfalls. To stabilize continual evolution, we adopt batch-level post-verification with rollback, which commits notebook edits only if they improve performance on the same batch, reducing noisy updates and preventing regressions. M2Note supports both self-evolving, where the same VLM acts as solver and supervisor, and cross-model evolving, where a stronger supervisor guides a weaker solver, enabling capability transfer without weight updates. Experiments on six multimodal reasoning benchmarks show consistent improvements across domains and backbones, while achieving strong cost and sample efficiency and remaining complementary to Chain-of-Thought (CoT) prompting.

Haiwen Li, Jing Tang, Rui Chen +2

AMAP, Alibaba Group

Memory-Augmented VLMs VLM Reasoning VLM Hallucination Mitigation Continual Learning for VLMs Multimodal Understanding

Jul 1, 2026cs.CV

H-OPD: Confidence Aware Heterogeneous Multi-Teacher Multimodal On-policy Distillation

On-policy distillation (OPD) has recently emerged as an effective post-training paradigm by providing supervision on student-generated trajectories. However, existing OPD methods for multimodal reasoning usually rely on a static teacher routing, assigning each sample to a single teacher based on modality or task type. This ignores that visual grounding and abstract reasoning may dominate different decoding steps, making a single teacher insufficient for the full trajectory. To this end, H-OPD is proposed as a confidence-aware heterogeneous multi-teacher OPD framework for multimodal reasoning. By verifying the complementarity of heterogeneous teachers in the same reasoning process, H-OPD replaces task or sample level teacher routing with token-level teacher arbitration along the shared student trajectory. H-OPD employs vision-to-language description transfer to enable text-only teachers to access key visual semantics, and uses a confidence-aware arbitration mechanism to dynamically combine vision-language teacher and text-only teachers at each token. Extensive evaluations over 11 widely-used reasoning benchmarks showcase the superior performance of our method.

Qixiang Yin, Huanjin Yao, Yuchen Cai +5

Beijing University of Posts and Telecommunications · Zhongguancun Academy, Beijing, China · ByteDance +3

Cross-Modal Knowledge Distillation VLM Distillation Multimodal Reasoning Multi-Teacher Knowledge Distillation On-Policy Distillation Multimodal Understanding

Jul 1, 2026cs.CV

Homer: Understanding Long-form Videos with Hierarchical Memory and Agentic Reasoning

Multimodal large language models excel on short clips but struggle on hour-long videos in an online setting, where frames are processed incrementally under limited memory. Existing online methods either retain compact visual representations that lack semantic structure, or build higher-level memory stores organized around temporal proximity rather than explicit causal links, leaving multi-hop narrative reasoning to be reconstructed by the LLM at every query. We bridge this gap with \textsc{Homer}, a Hierarchical Online Memory Exploration and Reasoning framework. \textsc{Homer}'s memory mirrors the multi-scale structure of long videos, ranging from raw perception, to recurring entities, to events connected by explicit temporal and causal relations. Its agentic reasoner then explores this memory the way humans do, locating the relevant scene, looking up details, and composing the answer through multi-round memory retrieval, with a harness that verifies and corrects each step. \textsc{Homer} outperforms the previous best agent method by

+5.5

,

+10.8

, and

+4.4

points on M3-Bench-robot, M3-Bench-web, and Video-MME-Long, and consistently lifts three various LLM backbones, indicating a model-agnostic structural capability for grounded retrieval over long videos.

Yixin Ji, Fanghua Ye, Juntao Li +5

1Soochow University · 2Tencent Hunyuan Multimodal Department

Long-Video Understanding Memory-Augmented Video Understanding Long-Video QA Memory-Augmented Language Model Agents Multimodal Understanding

Jul 1, 2026cs.CV

Multimodal Continuous Reasoning via Asymmetric Mutual Variational Learning

Multimodal Large Language Models (MLLMs) are often constrained by a language-space bottleneck, forcing complex visual reasoning into discrete tokens which can lose perceptual nuance. A promising alternative is continuous latent reasoning, where the goal is to discover implicit reasoning pathways that bridge the multimodal query and the final answer. However, this introduces a severe train-inference mismatch: a training-time posterior, conditioned on the ground-truth answer, can exploit answer-dependent shortcuts. Standard variational training then forces the inference-time prior to mimic a posterior that has access to information unavailable at test time, leading to poor performance. To address this, we propose Asymmetric Mutual Variational Learning (AMVL), a framework that resolves this mismatch via a bidirectional calibration objective. A forward KL divergence trains the target-agnostic prior to match the posterior, while a novel reverse KL divergence simultaneously regularizes the posterior, preventing it from collapsing into inference-incompatible regions and mitigating this ``answer leakage''. We provide theoretical analysis formalizing this leakage as prior contamination and prove that our dual-KL objective reduces it. We instantiate AMVL in a latent-integrated MLLM and show that it consistently outperforms strong discrete and latent-reasoning baselines, improving the average score on the complex BLINK benchmark by +10.83 and achieving gains of up to +32.00 on individual reasoning tasks, with analyses confirming improved latent-space stability.

Shijie Li, Yilin Gao, Siyuan Yang +7

Shanghai Jiao Tong University · Ant Group

Multimodal Large Language Models Latent Visual Reasoning Continuous Latent Reasoning Multimodal Foundation Models Variational Inference Latent Reasoning+1

Jul 1, 2026cs.CV

Wake up for Touch! Mask-isolated Tactile Alignment Learning in MLLMs

Touch supplies the physical grounding needed to perceive intrinsic material properties, such as friction and compliance, that vision alone often cannot resolve. Recent efforts for equipping multimodal LLMs with this tactile sense, however, expose a zero-sum trade-off: the limited parameter budget of compact models forces a choice between acquiring the new sensory modality and preserving the established vision-language reasoning. We present Splash, a mask-isolated tactile alignment learning framework for MLLMs. Splash quantifies the significance of each pretrained parameter, and partitions the parameter space into a dormant and critical subspace. While the frozen critical subspace acts as a stable anchor to safeguard general visual knowledge, Splash updates the isolated dormant subspace to internalize tactile alignment towards LLMs. This selective, non-destructive expansion effectively prevents catastrophic forgetting and ensures non-destructive modality expansion. Extensive experiments show that Splash effectively achieves tactile reasoning without additional inference overhead in the LLM part, demonstrating state-of-the-art performance on visuo-tactile benchmarks, including SSVTP, TVL, and TacQuad, while preserving its original general-purpose capabilities.

Yoonhyung Park, Minji Kim, Sungwon Moon +1

Division of Artificial Intelligence & Software, Ewha Womans University

Visuo-Tactile Representation Learning Multimodal Large Language Models Multimodal Foundation Models Continual Learning for VLMs Parameter-Efficient Fine-Tuning Multimodal Understanding

Jul 1, 2026cs.CV

Rosetta: Composable Native Multimodal Pretraining

Achieving true artificial general intelligence requires foundation models capable of integrating new modalities without forgetting prior knowledge. However, accommodating continuous generative objectives alongside discrete understanding tasks causes severe gradient conflicts. Existing architectures, including standard Mixture-of-Experts (MoE), are highly susceptible to representation overwriting. Even structurally partitioned paradigms like Mixture-of-Transformers (MoT) remain vulnerable to catastrophic forgetting, severely impeding multimodal scalability. In this work, we introduce Rosetta, a composable native multimodal pretraining framework designed for seamless and non-destructive modality expansion. Rosetta adopts a modular paradigm where core foundational knowledge is preserved within global shared experts, while modality-specific capabilities are distributed across plug-and-play experts. To guarantee non-destructive composition, we propose Momentum-Anchored Orthogonal Projection (MAOP). MAOP leverages the optimizer's momentum state as an implicit semantic anchor, selectively neutralizing conflicting gradient components from new modalities while preserving synergistic updates. Extensive evaluations demonstrate that, while standard MoE and MoT architectures suffer catastrophic forgetting of previously acquired knowledge, Rosetta robustly preserves established language and visual understanding. Furthermore, it delivers superior image generation and unlocks cross-modal synergy, paving the way for truly composable and unified multimodal foundation models. To facilitate further multimodal research, we release our code and checkpoints to the community. Project page at https://rosetta-lmm.github.io/.

Xiangyue Liu, Zijian Zhang, Miles Yang +3

HKUST · Tencent Hunyuan

Unified Multimodal Models Multimodal Pretraining Multimodal Generation Gradient Projection Multimodal Foundation Models Continual Learning for VLMs+2

Jun 30, 2026cs.CV

CoLT: Teaching Multi-Modal Models to Think with Chain of Latent Thoughts

Chain-of-thought (CoT) reasoning has enabled multi-modal large language models (MLLMs) to tackle complex visual reasoning tasks by generating explicit intermediate reasoning steps in natural language. However, this text-based reasoning paradigm is inherently slow at inference time with even thousands of tokens and fundamentally constrained by the expressiveness of natural language. In this paper, we propose CoLT, (Chain of Latent Thoughts), a novel framework that teaches multi-modal models to reason through a chain of latent thought representations instead of verbose text tokens, which can perform thinking with as few as 3 steps. Naively forcing the model to think with latent states easily produces meaningless semantics and makes training unstable. To effectively regulate the latent reasoning process, we introduce a lightweight external decoder that provides step-level supervision for each latent reasoning step in two complementary directions: a forward mode that decodes latent thoughts into the textual reasoning of the next step, and a backward mode that aligns decoder hidden states with the model's latent thoughts given preceding textual context. We further incorporate internal supervision that encourages coherent step-by-step latent transitions. The decoder and internal supervision are removed during inference to maintain high efficiency of latent reasoning. Extensive experiments on eight benchmarks demonstrate that CoLT not only outperforms existing latent reasoning methods such as CODI and SIM-CoT, but also surpasses latent visual reasoning approaches that rely on auxiliary images with costly annotation requirements. Compared to text CoT methods, CoLT can notably reduce the inference time by 10.1

\times

and text decoding time by 22.6

\times

. Code is released at https://github.com/hulianyuyy/CoLT.

Lianyu Hu, Shengqian Qin, Zeqin Liao +4

Nanyang Technological University, Singapore · Shanghai Jiao Tong University, China · NKIARI, Shenzhen Futian, China +2

Efficient Multimodal Inference Efficient Language Model Reasoning Latent Visual Reasoning Multimodal CoT Reasoning Multimodal Foundation Models Multimodal Understanding+1

Jun 30, 2026cs.CV

Attend, Transform, or Silence: Operator-Level Visual Skipping for Efficient Multimodal LLM Inference

Multimodal large language models (MLLMs) increasingly process long visual-token sequences, increasing the overall inference computation. Existing acceleration methods usually remove visual tokens or skip visual-token updates in entire layers, but these coarse strategies may discard fine-grained evidence or suppress useful operators together with redundant ones. In this paper, we study visual-token computation from an answer-observable perspective and find that late visual-token updates can remain large while having little effect on answer-token representations. Motivated by this answer-silent redundancy, we decompose each Transformer layer into attention and FFN operators and show that useful visual computation is often operator-dominant and layer-dependent. We propose an operator-level visual-token skipping framework that preserves the full visual-token sequence while selectively bypassing redundant attention, FFN, or both. Experiments across three MLLM architectures and 10 VQA benchmarks show that our method achieves strong efficiency-accuracy trade-offs, reducing \textbf{33.7%} TFLOPs on Qwen3-VL while retaining \textbf{99.5%} of the vanilla model performance.

Zhaoyang Luo, Runmin Dong, Miao Yang +4

Tsinghua Shenzhen International Graduate School, Shenzhen, China · Sun Yat-sen University, Zhuhai, China · Tsinghua University, Beijing, China +1

Efficient VLM Inference Multimodal Large Language Models Visual Token Pruning Multimodal Understanding

Jun 30, 2026cs.CV

Token-Sparse Medical Multimodal Reasoning via Dual-Stream Reinforcement Learning

Vision-language models (VLMs) combining reinforcement learning (RL) ignite remarkable progress in multimodal reasoning, yet still struggle with medical images, which typically exhibit extremely sparse visual evidence to inform clinical decision-making. We recognize that pruning visual tokens outside the grounding region greatly enhances medical reasoning. However, a united RL framework for active visual token pruning (VTP) and medical multimodal reasoning remains unestablished. Here, we propose a dual-stream RL framework, ViToS, to fulfill token pruning and question answering. ViToS trains one policy model with two task branches, where one focuses on grounding while the other conducts token-sparse reasoning after VTP. Furthermore, we solve the coupled policy learning problem by introducing the cross-feedback sequential optimization, avoiding gradient conflict and facilitating convergence of the shared policy model. Evaluated on seven medical benchmarks, our method reduces visual tokens to 77% of the original sequence length while achieving a 108.27% relative performance on Lingshu-7B and 104.16% relative performance on HuatuoGPT-Vision-7B. Overall, ViToS delivers superior performance and inference speedup, establishing an efficient paradigm for medical multimodal reasoning.

Kaitao Chen, Weiqian Zhao, Jiamin Wu +6

1Fudan University · 2Shanghai Artificial Intelligence Laboratory · 3Shanghai Jiao Tong University +1

Efficient Multimodal Inference Medical VQA Multimodal Reasoning Medical VLMs Visual Token Pruning RL for VLMs+1

Jun 30, 2026cs.CV

MS-Resampler: Multi-Scope Visual Resampling for Efficient Multimodal LLMs

Multimodal large language models (MLLMs) typically employ resampling-based projectors to transform dense visual features into a compact token sequence for language modeling. Most existing resamplers adopt a single, fixed aggregation scope via global cross-attention, which can blur fine-grained local evidence and limit the ability to capture both local details and global context within a fixed token budget. In this work, we propose MS-Resampler, a multi-scope visual resampling framework for MLLMs. MS-Resampler instantiates multiple scope-specific resamplers by injecting explicit spatial scope priors into the resampling attention, enabling each branch to aggregate visual information at a particular granularity from local to global. The outputs of these scope-specific resamplers are then adaptively fused to produce the final visual representations for language modeling. Extensive experiments on ten public multimodal benchmarks show that MS-Resampler consistently improves visual understanding and multimodal reasoning over conventional single-scope resamplers, while introducing only minimal computational overhead.

Zhongyang Li, Yaqian Li, Faming Fang +6

Li Auto Inc. · East China Normal University.

Multimodal Large Language Models Multimodal Fusion Visual Token Compression Multimodal Understanding

Jun 30, 2026cs.CV

Learning to Deny: Action Denial in Multimodal Large Language Models

Multimodal large language models (MLLMs) have rapidly advanced video understanding, achieving strong zero-shot and few-shot recognition across standard benchmarks. Yet their ability to deny an action by recognizing when an activity is not happening despite strong contextual cues remains largely unexplored. We introduce UCF101-AD, a large-scale benchmark consisting of paired Action-Presence and Action-Denial clips, designed to evaluate this capacity for denial. Each negative video in UCF101-AD preserves the same contextual and motion cues, including persons, objects, and locations, as its positive counterpart, but the defining action itself is explicitly absent. Evaluating 20 state-of-the-art MLLMs reveals a consistent failure: models that exceed 85% accuracy on the positive action classes collapse below 50% on their action-denial counterparts, indicating a strong inclination to affirm plausible actions rather than verify that they truly occur. This exposes a critical blind spot in modern video understanding: the inability to reason causally about whether a motion actually happens. To probe this issue, we explore a causal graph formulation, CausalAct, which expresses scene structure through natural-language prompts linking context, interaction, and motion. Incorporating such causal cues substantially reduces false positives, demonstrating that denial is a learnable reasoning skill. UCF101-AD provides a new lens for diagnosing and improving causal reasoning in multimodal models. Dataset and relevant code: https://github.com/raiyaan-abdullah/Learn-to-Deny.

Raiyaan Abdullah, Shehreen Azad, Yogesh Singh Rawat

Institute of Artificial Intelligence, University of Central Florida, USA

Multimodal Large Language Models Causal Reasoning in Videos Multimodal Model Evaluation Multimodal Understanding