JARVIS-1: Open-World Multi-task Agents with Memory-Augmented Multimodal Language Models.
Zihao WangShaofei CaiAnji LiuYonggang JinJinbing HouBowei ZhangHaowei LinZhaofeng HeZilong ZhengYaodong YangXiaojian MaYitao LiangPublished in: CoRR (2023)
Keyphrases
- language model
- multi task
- open world
- incomplete information
- multi task learning
- language modeling
- learning tasks
- multi agent
- multi agent systems
- closed world
- n gram
- probabilistic model
- information retrieval
- multiagent systems
- autonomous agents
- multi class
- feature selection
- learning problems
- multi modal
- transfer learning
- integrity constraints
- data exchange
- dynamic environments
- knowledge base
- keywords
- databases