Definition
Agent memory is the persistent state an AI agent carries across turns, sessions, or tasks — everything beyond what fits in the current context window. Designs span scratchpads, vector stores, structured knowledge bases, and explicit episodic memories, all wrestling with the same tension: keep enough to be useful, prune enough to stay coherent.
Episodes covering this
Worth reading next
Papers we haven't done a deep dive on yet, but would recommend on this topic.
- LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory
- LoCoMo: Long-Context Modular Memory for Dialogue State Tracking
- Cognitive Architectures for Language Agents
- ExpeL: LLM Agents Are Experiential Learners
- MemGPT: Towards LLMs as Operating Systems
- A-MEM: Agentic Memory for LLM Agents
- Generative Agents: Interactive Simulacra of Human Behavior
- Larimar: Large Language Models with Episodic Memory Control
- ReadAgent: A System for Agent-Based Long-Context Reading Tasks
- Voyager: An Open-Ended Embodied Agent with Large Language Models
- Reflexion: Language Agents with Verbal Reinforcement Learning