CORTEXA
← Browse
arxivcs.CVcs.CL2026-07-01

Latent Visual Cache for Video Reasoning

Yongheng Zhang, Zhipeng Xu, Hao Wu, Yinghui Li, Di Yin, Xing Sun, Philip S. Yu

Video reasoning requires Large Multimodal Models (LMMs) to remain grounded in dense evidence, yet existing systems largely adopt "read-once, generate-many" paradigm, in which visual grounding weakens during generation. This phenomenon has been widely observed and is known as Visual Anchoring Decay. To fill this gap, we introduce Latent Video Cache (Latent-VC), a recurrent latent visual cache inserted into the decoder to preserve compact visual memories throughout reasoning. The cache is trained with supervised contrastive cache alignment and vision-grounded GRPO with a latent grounding reward, while maintaining strict train-inference alignment through native decoder hidden states. Built on Qwen3.5-9B, Latent-VC consistently outperforms strong CoT and SFT+GRPO baselines across six video benchmarks, with especially clear gains on grounding-intensive and long-video tasks. In addition, it also achieves higher accuracy with substantially shorter responses, suggesting that latent visual caching improves video reasoning by preserving visual evidence rather than relying on longer textual chains.

View free PDFSource page

Related papers

arxivcs.CVcs.AIcs.CL2026-06-26

Animation2Code: Evaluating Temporal Visual Reasoning in Video-to-Code Generation

Anya Ji, Abhijith Varma Mudunuri, David M. Chan, Alane Suhr

While recent vision-language models (VLMs) have achieved significant improvements on static visual-to-code tasks such as generating code for webpages, charts, or SVGs, it remains unclear whether they can recover temporal dynamics when motion is present. To this end, we introduce…

View free PDFSource page
arxivcs.CVcs.CL2026-07-03

ProLaViT: Learning Progressive Latent Visual Thoughts in Structured Latent Space

Peiming Li, Yifan Wang, Xiaotian Zhang, Zhiyuan Hu, Shiyu Li, Zheng Wei, et al.

Multimodal Large Language Models (MLLMs) have achieved remarkable progress but still struggle with complex visual reasoning tasks requiring multi-step perception and logical deduction. While explicit visual generation incurs prohibitive computational costs, existing latent approa…

View free PDFSource page
arxivcs.CVcs.AIcs.CL2026-07-20

Thinking in Video: Can Video Generators Really Reason About the Real World?

Yongheng Zhang, Guang Yang, Ruihan Hou, Qiguang Chen, Ziang Liu, Xiaolong Liu, et al.

Recent advances in world models and video generation have given rise to an emerging reasoning paradigm that leverages video generative models to simulate, predict, and reason about real-world dynamics. We redefine this paradigm as Thinking in Video, where video is not merely an o…

View free PDFSource page
arxivcs.CVcs.CL2026-07-15

SD-MAR: Multi-image Analytical Reasoning via Synthetic Data and Reinforcement Learning

Shiyu Yuan, Sourav Sanjukta Bhabesh, Zhe Wang, Dmitriy Bespalov, Wesley Rose, Huzefa Rangwala

Vision Language Models (VLMs) demonstrate strong perceptual abilities but remain limited in tasks requiring analytical reasoning across multiple visual states, such as multi-image comparison, change detection, and multi-step visual inference. These capabilities are critical for r…

View free PDFSource page
arxivcs.CVcs.AIcs.CL2026-07-01

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

Yixin Ji, Fanghua Ye, Juntao Li, Bo Zhao, Zexuan Qiu, Zhaopeng Tu, et al.

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…

View free PDFSource page
arxivcs.CVcs.AIcs.CLcs.MA2026-07-20

O-VAD: Industrial Video Anomaly Detection through Object-Centric Tracking and Reasoning

Mei Yuan, Qi Long, Qifeng Wu, Zhenyang Li, Yizhou Zhao, Lei Wang, et al.

Industrial Video Anomaly Detection (IVAD) aims to identify anomalous objects and events in an industrial process, which is crucial for modern manufacturing and quality control systems. Existing VLM-based anomaly reasoning methods are capable of detecting open-ended anomalies in g…

View free PDFSource page