AI.info

Sibei Yang

Explore Sibei Yang on AI.info.

• WeaveTime: Stream from Earlier Frames into Emergent Memory in VideoLLMs
• RefAny3D: 3D Asset-Referenced Diffusion Models for Image Generation
• Intervene-All-Paths: Unified Mitigation of LVLM Hallucinations across Alignment Formats
• Why LVLMs Are More Prone to Hallucinations in Longer Responses: The Role of Context
• Augmenting Moment Retrieval: Zero-Dependency Two-Stage Learning
• Closed-Loop Transfer for Weakly-supervised Affordance Grounding
• Eyes Wide Open: Ego Proactive Video-LLM for Streaming Video