AgentCVR
AgentCVR: Active Multi-Agent Cross-Video Reasoning via Script-Simulated Reinforcement Learning
Yilun Qiu, Jiahe Wang, Cilin Yan, Jiayin Cai, Xiaolong Jiang, Yao Hu, et al. · 2026-05-28
- Management
- T1 Static Memory
- Memory is built once and reused without change during deployment.
- Representation
- Source- and modality-specific
How experience becomes memory
Each path is one memory component, traced from what is stored to how it is found again.
cross-video audiovisual evidence
- Represented asRetained perceptual source evidenceSource- and modality-specific
- Stored inHierarchical multi-tier store
- Retrieved byPlanner / tool-assisted retrieval, Metadata / spatiotemporal retrieval
- Source evidenceSource directly available
A Master Agent coordinates visual and audio agents to acquire targeted evidence across multiple videos.Full-paper method / memory representation and retrieval · PDF p. 2 · Figure 1
Related systems
- UniversalRAGT1 · Video & streaming
- Anim-DirectorT1 · Video & streaming
- Deep Video DiscoveryT1 · Video & streaming
- EgoButlerT1 · Video & streaming
- EvoVidT1 · Video & streaming
- HD-EPIC Long-Video ReasoningT1 · Video & streaming