Towards Autonomous Memory Agents
This addresses the limitation of reactive memory growth in AI agents, offering a more efficient and proactive approach for enhancing LLM performance, though it appears incremental by building on existing memory agent frameworks.
The paper tackles the problem of passive memory agents in LLMs by proposing autonomous memory agents that actively acquire, validate, and curate knowledge, resulting in improvements such as a 14.6-point gain on HotpotQA and a 7.33-point gain on AIME25.
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.