PreCoMem: Predictive Cognitive Memory for Self-Evolving Long-Term Dialogue Agents
We introduce PreCoMem, a memory consolidation framework inspired by predictive coding that represents memory as confidence-weighted beliefs about the user's state.
ProofPaper ↗
Key points
- Long-horizon dialogue agents need to keep user memories stable enough for personalization while adapting when the user genuinely changes.
- Existing memory systems often append every observation or overwrite old entries by heuristic rules, which makes transient noise hard to distinguish from real belief drift.
- According to effective surprisal, a two-threshold gate maps each turn to one of three deterministic updates: MAINTAIN reinforces supported beliefs, PROFILE stores ambiguous cues as low-confidence hypotheses, and CORRECT softly decays reliably contradicted facts.
- Experiments on LoCoMo, LongMemEval-S, and PersonaMem-v2 show consistent state-of-the-art accuracy; our new DiMoBench confirms that PreCoMem adapts to non-stationary belief drift without over-reacting to noise.
Sources (1)
- [1]PreCoMem: Predictive Cognitive Memory for Self-Evolving Long-Term Dialogue AgentsarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 21, 09:21 AM
We introduce PreCoMem, a memory consolidation framework inspired by predictive coding that represents memory as confidence-weighted beliefs about the user's state.
Long-horizon dialogue agents need to keep user memories stable enough for personalization while adapting when the user genuinely changes.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 20, 2026How Eon built multi-cloud investigation with an MCP Gateway on Amazon Bedrock AgentCore
- Oct 20, 2026Memory as Code: Declarative Retrieval over Codified Long-Term Memory
- Oct 20, 2026Do Diffusion LLMs Scale? Testing Mercury 3 + Decide
- Oct 20, 2026Cache Engineering for Agents
- Oct 20, 2026Evolving Agent Interfaces
- Oct 20, 2026Automating ML for Science: Can Frontier Agents Climb Scientific Hills in the Wild?