AgentPoison: Red-teaming LLM Agents via Poisoning Memory or Knowledge Bases.
Zhaorun ChenZhen XiangChaowei XiaoDawn SongBo LiPublished in: CoRR (2024)
Keyphrases
- knowledge base
- multi agent systems
- reasoning process
- multi agent
- multiple agents
- autonomous agents
- multiagent systems
- software agents
- intelligent agents
- cooperative
- past experience
- knowledge representation
- mobile agents
- dynamic environments
- artificial agents
- knowledge based systems
- knowledge acquisition
- distributed agents
- abductive logic programming
- interacting agents
- description logics
- learning agents
- knowledge sources
- expert systems
- agent systems
- single agent
- game theoretic
- decision making
- dynamically created
- memory usage
- engineering design
- memory requirements
- resource allocation
- wordnet
- logic programs
- artificial intelligence
- agent model
- agent architecture
- memory space
- main memory
- agent behavior
- website