DeepYardDeepYard
M

MemMA

Multi-agent memory framework with self-evolution for LLM agents

Open SourceFree

About

Research framework that enhances LLM agent memory through multi-agent coordination and in-situ self-evolution. Tackles strategic blindness in memory systems by implementing explicit reasoning chains and feedback loops for both memory construction and retrieval. Designed for researchers and developers building sophisticated agent systems that require robust long-term memory management.

Details

Type
Integrations
Language

Tags

memorymulti-agentframeworkopen-sourceautonomousrag