MedGraph introduces graph memory for medical dialogue
MedGraph uses graph memory to maintain patient state, outperforming baselines by 10 points on MedMemoryBench.
Stories per day · 30 days
2 RSS
Relatedgraph-memory1llm-agents1medical-dialogue1ontology1personal-assistants1reinforcement-learning1
4 stories, newest first · one card per story
Stories tagged #memory-systems
EvoMemState framework evolves memory states for LLM personal agents
EvoMemState augments facts with relational and reasoning states to improve personalized assistance in agents.
Fortunate Recall: Ontology-Driven Memory Lifecycle Management for Persistent Coherence in LLMs
Fortunate Recall introduces ontology-based lifecycle policies for LLM memory, cutting confabulation roughly in half (e.g., 45.1% to 22.4%) versus Mem0.
Revoked but Still Authoritative: An Empirical Study of Revocation Enforcement in Agent-Memory Systems
An empirical study finds no major agent-memory system enforces fact revocation at retrieval, causing agents to act on superseded, unsafe information.