MemPO: Self-Memory Policy Optimization Algorithm Improves Long-Horizon Agent Performance While Reducing Token Usage
By
[Submitted on 28 Feb 2026 (v1), last revised 15 Jun 2026 (this version, v4)]
Crusty in the right places. Worth the chew.
Summary
This paper introduces MemPO (Self-Memory Policy Optimization), a novel algorithm that enables long-horizon AI agents to autonomously summarize and manage their own memory during environment interactions, rather than relying on external memory modules. The approach improves credit assignment based on memory effectiveness, allowing the model to selectively retain crucial information while significantly reducing token consumption. Experimental results show MemPO achieves absolute F1 score gains of 25.98 over the base model and 7.1 over the previous state-of-the-art baseline, while reducing token usage by 67.58% and 73.12%.
Key quotes
· 3 pulledLong-horizon agents face the challenge of growing context size during interaction with environment, which degrades the performance and stability.
Existing methods typically introduce the external memory module and look up the relevant information from the stored memory, which prevents the model itself from proactively managing its memory content and aligning with the agent's overarching task objectives.
MemPO achieves absolute F1 score gains of 25.98 over the base model and 7.1 over the previous SOTA baseline, while reducing token usage by 67.58% and 73.12%.
You might also wanna read
δ-mem: A Compact Online Memory Mechanism for Efficient Long-Context LLM Processing
The article presents δ-mem, a lightweight memory mechanism for large language models that augments frozen full-attention backbones with a co
arxiv.org·1mo ago
Memori launches agent-native persistent memory infrastructure using structured knowledge graphs from agent trace data
Memori is a new agent-native memory infrastructure that enables AI agents to create structured, long-term persistent memory directly from ag
Product Hunt·1mo ago
Robot Memory System Enables AI Robots to Learn from Past Experiences
Robot Memory (robotmem) is a persistent memory system for AI robots that enables robots to learn from past experiences. The system stores ep
github.com·3mo ago
CogniMemo: AI Memory Platform for Long-Term Context and Learning
CogniMemo is an AI memory platform that provides long-term memory capabilities for AI systems. It enables AI to remember users, preferences,
Product Hunt·6mo ago
AgentMemory: Open-source persistent memory tool for AI coding agents
AgentMemory is an open-source tool that gives AI coding agents (like Claude Code, Codex, Cursor, etc.) persistent memory across sessions, so
Product Hunt·1mo ago
YourMemory: Open-source persistent memory layer for AI agents using Ebbinghaus forgetting curve
YourMemory is an open-source persistent memory layer for AI agents that implements Ebbinghaus forgetting curve decay to mimic human memory.
github.com·1mo ago