InferAct: Inferring Safe Actions for LLM-Based Agents Through Preemptive Evaluation and Human Feedback.
Haishuo FangXiaodan ZhuIryna GurevychPublished in: CoRR (2024)
Keyphrases
- multiple agents
- artificial agents
- agent behavior
- action selection
- human operators
- multi agent
- multi agent systems
- autonomous agents
- software agents
- human activities
- decision theoretic
- explicitly represent
- cooperative
- human behavior
- resource allocation
- intelligent agents
- multiagent systems
- mobile agents
- interacting agents
- scheduling problem
- assessment tool
- intelligent behavior
- dynamic environments
- agent based simulation
- human teacher
- agent interactions
- rational agents
- state transitions
- agent teams
- agent receives
- single machine
- cooperative multiagent systems
- reasoning about actions
- personality traits
- human subjects
- bdi agents
- past experience
- reasoning process
- agent architecture