Streaming video LLMs must retain evidence before its relevance to future tasks is known and respond when sufficient evidence becomes available. The challenge is to form reusable factual memory without compromising real-time perception. We introduce OneStreamer, which jointly learns query-independent evidence recording...
Xiang-Yun Zeng, Yuan-Dong Yang, Zhi-Qiu Zhang et al.· 1 citation
Visual agents must turn video observations into timely decisions to track progress and choose what to do next. Building on the emerging System One paradigm, we introduce VideoJev, a family of models that cast video understanding and interaction as structured decision tasks. We adapt pretrained multimodal large language...
Yu-Han Zhu, Xiang-Yu Zeng, Qing-Yi Si et al.· 0 citations
A retention-aware review mechanism that tracks mastered prompts and periodically reintroduces them to periodically reintroduces them to the model of previous solutions, demonstrating robust generalizability across modalities and algorithms.
Chuan-Yu Qin, Chenxu Yang, Qing-Yi Si et al.· arXiv.org· 1 citation
Across six real-world scenarios, human raters prefer JoyAI-VL-Interaction over the in-app video-call assistants of Doubao and Gemini by a wide margin, and the first open, vision-driven interaction model released together with its training recipe, data, and complete deployable system.
Ding-Yu Yao, Jun Zhou, Chenxu Yang et al.· arXiv.org· 10 citations
CompassOPD is proposed, which removes this offset and transfers the within-family shift, while a frozen student reference anchors updates to the student's initial policy, and shows that CompassOPD consistently outperforms standard cross-family OPD.
Nai-Bin Gu, Qing-Yi Si, Chen-Xu Yang et al.· 0 citations
This work presents FORGE (Fabricated Orchestrated Reasoning chain for aGent Exploitation), a two-level attack that combines intra-document reasoning fabrication with inter-document chain coordination to hijack subtask planning.
Yu-Ru Pan, Ziheng Zhang, Junxiang Lei et al.· arXiv.org· 1 citation
Results show that EMAS can turn experience from new samples into reusable updates to MAS topology and prompts, and is best or tied in six of eight model--benchmark settings.
Chao Fei, Qing-Yi Si, Kai-Huan Liang et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
