The paper introduces EvidenceConnect, a data‑synthesis framework that generates training examples by zooming into related image regions to create questions requiring comparison, aggregation, or chaining of visual evidence. By retaining only reference answers agreed upon by independent vision‑language models and discarding questions that survive region ablation, the authors train base VLMs with RLVR to produce Zooming without Zooming v2 (ZwZ2). Although the synthesized data are not benchmark‑specific, ZwZ2 shows improvements across fine‑grained and spatial perception, embodied spatial reasoning, multimodal long‑context understanding, multimodal agent tasks, and general visual reasoning, demonstrating that strengthening the shared capability of connecting evidence yields more generalizable multimodal perception.
By Lai Wei, Yatai Li, Liangbo He, Jun Lan, Renjie Gu, Linhan Cao, Junkai Lin, Huijia Zhu, Weiqiang Wang, Linghe Kong, Yue Wang, Zhuosheng Zhang, Weiran Huang
VideoLVR introduces a method for multimodal large language models to maintain visual grounding during chain-of-thought reasoning by constructing latent visual evidence from local video regions at relevant moments. It uses a visual token prediction objective that unifies visual evidence reconstruction and text generation, allowing models to interleave text tokens with latent visual states. The approach is further extended with reinforcement learning, treating visual tokens as actions and optimizing policies with final-answer rewards, thereby improving video understanding without expanding the multimodal context.
By Shilong Li, Xingyuan Bu, Hao Shi, Yun Xing, Weixin Feng, Qiangpeng Yang, Shilei Wen
arXiv:2506.09557v2 Announce Type: replace-cross
Abstract: While Multimodal Large Language Models (MLLMs) demonstrate impressive performance in benign scenarios, their cognitive reliability deteriorat...
By Zhaoyang Wei, Bowen Jiang, Xumeng Han, Jiashu Li, Xuehui Yu, Yuling Liu, Guorong Li, Zhenjun Han, Jianbin Jiao
arXiv:2608. 10954v1 Announce Type: cross Abstract: While Multimodal Large Language Models (MLLMs) demonstrate impressive performance in benign scenarios, their cognitive reliability deteriorates significantly in complex scenes under adverse conditions.
By Zhaoyang Wei, Bowen Jiang, Xumeng Han, Jiashu Li, Xuehui Yu, Yuling Liu, Guorong Li, Zhenjun Han, Jianbin Jiao
The paper introduces ViMoD, a lightweight framework that dynamically manages visual memory for multimodal large language models. ViMoD uses Deformable Aggregation of Region-wise Tokens (DART) to create content‑adaptive coarse representations linked to recoverable fine tokens, and Temporal Routing for Adaptive Contextual Evidence (TRACE) to anticipate and adjust evidence needs during reasoning. On the Qwen3‑VL‑4B model, ViMoD achieves superior performance across eight reasoning benchmarks with only a 20% visual token budget and minimal additional parameters.
By Yicheng Xue, Han Wu, Jufeng Yang, Minjing Dong, Xinghao Chen, Hanting Chen, Jianyuan Guo
arXiv:2610.27341v1 Announce Type: new
Abstract: Vision-language models (VLMs) have advanced rapidly in multimodal reasoning, yet they remain limited on tasks requiring fine-grained understanding of s...
By Zerui Chen, Changrui Chen, Fei Ni, Jiankang Deng
arXiv:2610.02117v1 Announce Type: cross
Abstract: On-policy self-distillation has recently emerged as an effective approach for improving language-model reasoning by supervising students with a froze...
By Sophia Sirko-Galouchenko, Monika Wysoczanska, Andrei Bursuc, Nicolas Thome, Spyros Gidaris
The paper introduces TerraAgent, a training‑free framework that enhances remote sensing visual question answering by integrating three modules: GDR for demonstration retrieval, MACE for evidence‑typed sub‑question decomposition, and CVA for counterfactual evidence auditing. It demonstrates significant improvements on multiple benchmarks, raising accuracy, visual‑dependency scores, and reducing prior‑only answers and hallucinations without fine‑tuning.
By Jingwen Huang, Yiming Fang, Hanxiao Wu, Boyuan He, Keyi Luo
The paper introduces Selective Probability Mass Concentration (sPMC), a training framework that strengthens implicit visual grounding in multimodal large language models by selectively regularizing attention heads most responsive to visual evidence. sPMC treats attention over visual tokens as a spatial probability distribution and encourages mass to concentrate on semantically relevant regions using segmentation-derived priors, while leaving other heads unconstrained. Across six multimodal benchmarks, sPMC yields an average zero‑shot improvement of 3% and gains up to 11.3% for various models by regularizing only 3%–15% of their attention heads.
By Jiaqi Deng, Zonghan Wu, Zhan Heng, Xiaoshui Huang, Huan Huo, Guandong Xu
The paper introduces CAVE, a structured credit assignment method that enhances Vision‑Language Models by evaluating intermediate reasoning steps through belief updates, evidence acquisition, and adaptive focus control. CAVE is built on a GRPO framework and guides models to optimize each action, improving reliability in fragmented visual reasoning. The authors also present TRACER‑Bench, a new benchmark covering nonlocal and semantically confusable reasoning dimensions, and show that CAVE boosts performance on both public datasets and this benchmark while maintaining strong general multimodal results.
By Tengda Guo, Jie Leng, Hanlei Li, Yaoyuan Liang, Qingyue Zhang, Dian Yang, Mingyu Zhang, Yuhua Fu, Shao-Lun Huang
V‑Retrver is an evidence‑driven retrieval framework that treats universal multimodal retrieval as an agentic reasoning process grounded in visual inspection. It allows multimodal large language models to selectively acquire visual evidence through external tools, alternating between hypothesis generation and targeted visual verification. The approach is trained with a curriculum that blends supervised activation, rejection‑based refinement, and reinforcement learning, achieving an average 23.0% improvement in retrieval accuracy across multiple benchmarks.
By Dongyang Chen, Chaoyang Wang, Dezhao Su, Xi Xiao, Zeyu Zhang, Jing Xiong, Qing Li, Yuzhang Shang, Shichao Kan
MERIT (Multimodal Evidence Reasoning over Image Tables) is a three‑stage framework designed to enable language models to reason over table images, which preserve visual cues and structural details that serialized textual tables often lose. The approach first adapts a table‑specific encoder through reconstruction, localization, and color‑conditioned supervision to address visual evidence acquisition. It then fine‑tunes the model with supervised evidence selection and reasoning, followed by reinforcement learning that rewards evidence correctness, calculation, final answers, and output format, thereby improving retrieval and computational accuracy in table question answering and fact verification tasks.
By Haoyang Jiang, Zhuotong Fan, Mingyang Yi, Ju Fan, Lanqing Hu, Xiao Zan