Towards Autonomous Memory Agents

TL;DR

This paper introduces autonomous memory agents that actively acquire, validate, and curate knowledge to enhance language models, outperforming previous passive memory solutions and RL-based methods on multiple benchmarks.

Contribution

The paper presents U-Mem, a novel autonomous memory agent framework with cost-aware knowledge extraction and semantic-aware exploration strategies.

Findings

U-Mem outperforms prior memory baselines on HotpotQA and AIME25.

It surpasses RL-based optimization in memory management.

Demonstrates significant improvements in benchmark tasks.

Abstract

Recent memory agents improve LLMs by extracting experiences and conversation history into an external storage. This enables low-overhead context assembly and online memory update without expensive LLM training. However, existing solutions remain passive and reactive; memory growth is bounded by information that happens to be available, while memory agents seldom seek external inputs in uncertainties. We propose autonomous memory agents that actively acquire, validate, and curate knowledge at a minimum cost. U-Mem materializes this idea via (i) a cost-aware knowledge-extraction cascade that escalates from cheap self/teacher signals to tool-verified research and, only when needed, expert feedback, and (ii) semantic-aware Thompson sampling to balance exploration and exploitation over memories and mitigate cold-start bias. On both verifiable and non-verifiable benchmarks, U-Mem consistently beats prior memory baselines and can surpass RL-based optimization, improving HotpotQA (Qwen2.5-7B) by 14.6 points and AIME25 (Gemini-2.5-flash) by 7.33 points.