EgoMemReason is a new benchmark for week‑long egocentric video understanding that focuses on memory‑driven reasoning rather than simple perception tasks. It tests three memory types—entity, event, and behavior—across 500 questions, each requiring evidence from an average of 5.1 video segments and 25.9 hours of backtracking. Evaluation of 17 models shows that even the best achieves only 39.6% accuracy, highlighting the difficulty of long‑horizon memory in multimodal systems.
By Ziyang Wang, Yue Zhang, Shoubin Yu, Ce Zhang, Zengqi Zhao, Jaehong Yoon, Hyunji Lee, Gedas Bertasius, Mohit Bansal
arXiv:2610.17540v1 Announce Type: new
Abstract: Textual reasoning before acting has improved vision-language-action (VLA) models, but at the cost of latency: every action conditioned on the reasoning...
By Jack Lu, Yurong You, Boyi Li, Yulong Cao, Thomas Tian, Chaowei Xiao, Xiangyu Chen, Wenjie Luo, Boris Ivanovic, Marco Pavone, Yan Wang
arXiv:2606. 17924v1 Announce Type: cross Abstract: Current Vision-Language-Action (VLA) models face a trade-off between efficient action generation and explicit deliberation.
By Bochen Yang, Lianlei Shan
The paper introduces 4DMemory, a benchmark for long‑horizon spatiotemporal memory that includes the LongVideo dataset—about 900 hours of video with over 5,000 manually reviewed questions—and the 4DMemoryBench. It also presents the 4DMemory Harness, a system that combines four complementary skills, more than 22 specialized tool models, and a retrieval‑augmented generation component. Evaluation of 13 leading models shows that standalone foundation models perform poorly on long‑horizon tasks, while the 4DMemory Harness boosts performance by over 50% and reveals a performance drop of more than 10% compared to previous benchmarks.
By Yanhao Jia, Yi Zhu, Yiran Zhong, Weigao Sun, Feida Zhu, Pengkun Jiao, Sihan Yang, Ming Ma, Steven Hoi, Erik Cambria
arXiv:2606. 07512v1 Announce Type: cross Abstract: Current Vision-Language Models struggle with hours-long videos because processing full-length visual sequences induces prohibitive token explosion and attention dilution.
By Cong Chen, Guo Gan, Kaixiang Ji, ChaoYang Zhang, Zhen Yang, Guangming Yao, Hao Chen, Jingdong Chen, Yi Yuan, Chunhua Shen
ARGOS is a new benchmark and agent framework for multi‑camera person search that transforms the task from one‑shot retrieval into interactive reasoning using partial witness clues. The framework requires agents to plan questions, use spatial or temporal tools, and interpret ambiguous natural‑language responses within a limited turn budget, leveraging a Spatio‑Temporal Topology Graph that encodes camera connectivity and transition times. The benchmark includes 2,691 tasks across 14 real‑world scenarios, divided into semantic, spatial, and temporal tracks, and introduces Turn‑Weighted Success (TWS) as a metric that jointly measures correctness and turn efficiency, with current best agents achieving TWS scores of 0.383 and 0.590 on the spatial and temporal tracks respectively.
By Myungchul Kim, Kwanyong Park, Junmo Kim, In So Kweon
The paper introduces Reasoning and Flow Memory (FLOWMEM), a vision‑language‑action model that reuses successful latent reasoning steps instead of discarding them after each policy query. FLOWMEM dynamically retrieves and recomposes compatible latent fragments to form a reasoning route that aligns with the temporal structure of prior successful computations, then refines this route with current visual and proprioceptive data before generating actions. Experiments on RoboMME and LIBERO‑Plus show that FLOWMEM improves success rates by 1.7 and 4.1 percentage points over memory‑free policies, demonstrating the benefit of reusing latent computation for closed‑loop control.
By Hongyu Shi, Sen Zhao, Zuyu Zhang, Lifeng Shen, Ding Zou, Xinyu He, Xu Zhang, Qinghua Zhang
The paper introduces Latent Transition Memory (LTM), a method that stores interaction history as compact latent transitions between consecutive observations for vision‑language‑action models. Implemented as LTM‑π on the π₀.₅ architecture, it uses a frozen latent action encoder, a lightweight adapter, and causal KV caching to enable incremental inference without external reasoning models or memory‑specific annotations. Experiments on RoboMME and LIBERO‑Mem show that LTM‑π achieves state‑of‑the‑art success rates (58.54% and 80.5% respectively), and scaling the dataset or applying a 200‑token memory budget still yields high performance, demonstrating the method’s scalability and efficiency in real‑world robotic tasks.
By Qixin Sun, WenLong Gao, Xiansheng Chen, Zhongyuan Wang, Si Liu, Pengwei Wang
arXiv:2608.22869v1 Announce Type: cross
Abstract: While Vision-Language-Action (VLA) models have leveraged internet-scale pretraining and task-focused finetuning to achieve strong performance on long...
By Lars Osterberg, Maggie Wang, Mac Schwager
arXiv:2512. 05277v3 Announce Type: replace-cross Abstract: Vision-Language Models (VLMs) are increasingly deployed as the perception and reasoning backbone of autonomous agents acting in the wild, with autonomous driving (AD) being one of the most safety-critical instances.
By Kevin Cannons, Saeed Ranjbar Alvar, Mohammad Asiful Hossain, Ahmad Rezaei, Mohsen Gholami, Alireza Heidarikhazaei, Zhou Weimin, Yong Zhang, Mohammad Akbari
arXiv:2606.20092v3 Announce Type: replace
Abstract: Memory remains a critical bottleneck for long-horizon robotic manipulation, as standard Vision-Language-Action (VLA) policies often fail when task-...
By Ganlin Yang, Zhangzheng Tu, Yuqiang Yang, Sitong Mao, Junyi Dong, Tianxing Chen, Jiaqi Peng, Jing Xiong, Jiafei Cao, Jifeng Dai, Wengang Zhou, Yao Mu, Tai Wang
CoWorld Reasoner is an agentic framework that leverages a Video World Model (VWM) as a unified visual reasoning engine for Vision‑Language Models (VLMs). It exposes multiple visual operations through tools and skills, allowing the VLM to decide when to imagine, select an operation, generate visual content, and interpret the resulting evidence. The framework includes a VLM‑VWM Harness Co‑Evolution mechanism that jointly refines tools, skills, and execution code based on feedback, while keeping model weights fixed. Across seven benchmarks covering spatial, temporal, causal, and perceptual reasoning, CoWorld Reasoner outperforms direct VLM reasoning and improves accuracy on all benchmarks, achieving notable gains on MMSI and ERQA.
By Dongjie Cheng, Ruifeng Yuan, Ran Wei, Ning Song, Yongqi Li, Wenjie Li