MoTE: Mixture of Task Experts for Multi-Task Video Understanding

  • 类型:arxiv
  • 标识:2608.24763
  • 链接:https://arxiv.org/abs/2608.24763
  • 主分类:multimodal
  • 形态:method
  • TLDR:Procedural video-language models must solve heterogeneous tasks from the same visual evidence, including action recognition, forecasting, and procedure prediction. Dense transformer decoders share the same feed-forward networks across tasks, which can entangle task behavior and make controlled capability expansion difficult. Sparse Mixture-of-Experts (MoE) decoders provide conditional computation, but token-level learned routing is not naturally aligned with task-level procedural objectives. We propose MoTE (Mixture of Task Experts), a decoder architecture that converts large language model fe
  • 待LLM分类:否
  • 来源文件
  • /inbox/tom/_candidates/2026-08-27-agent-rag-longcontext-candidates.json