Research map / Architectures and efficiency / Attention and transformers
MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory Agent
Hongli Yu, Tinghong Chen, Jiangtao Feng, Jiangjie Chen, Weinan Dai, Qiying Yu, Ya-Qin Zhang, Wei-Ying Ma, Jingjing Liu, Mingxuan Wang, Hao Zhou
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| train memory agent with multi-conv DAPO RL instead of using MemAgent workflow untrained (ablation) | MemAgent · RULER-HQA and LongBench-QA | accuracy | Improved |
| apply MemAgent workflow without RL (ablation) | MemAgent · LongBench-QA | accuracy | No clear effect |
| MemAgent without RL versus backbone (ablation) | MemAgent · RULER-HQA | accuracy | Mixed |
| process text in chunks with fixed-length overwrite memory (RL-MemAgent-14B) | Qwen2.5-14B-Instruct · RULER-HQA 3.5M | accuracy: 80.47 (7K) -> 71.09 (3.5M) | Improved |
| process text in chunks with memory vs Qwen2.5-Instruct-1M | Qwen2.5-Instruct-14B-1M · RULER-HQA 896K | accuracy: 0.00 -> 75.78 | Improved |
| RL-MemAgent on LongBench-QA vs larger long-context models | Qwen2.5-14B-Instruct · LongBench-QA average | score: 50.7 (QwenLong-L1-32B) -> 51.0 | Improved |
| RL-MemAgent for summarization | Qwen2.5-14B-Instruct · LongBench-SUM GovReport | ROUGE average: 19.34 -> 21.80 | Improved |
| larger memory (4096) vs default memory 1024 (ablation) | MemAgent-14B · LongBench-QA | average score: 51.0 (m1024) vs 48.5 (m4096) | No clear effect |
| memory length 256 to 4096 on NIAH (ablation) | MemAgent · NIAH | accuracy: 97.92 to 98.18 at 512K (14B) | No clear effect |
| shift key information position in context (probe) (ablation) | MemAgent-14B · RULER-HQA probe | accuracy: 0.00 to 2.79 avg diff vs random | No clear effect |
| use RL-trained memory instead of RAG retrieval (top-K) | Qwen2.5-14B · RULER-HQA 3.5M | accuracy: 66.41 -> 71.09 | Improved |
Closest papers by meaning
- A Human-Inspired Reading Agent with Gist Memory of Very Long Contexts (ICML 2024 · similarity 0.88)
- Evaluating Memory in LLM Agents via Incremental Multi-Turn Interactions (ICLR 2026 · similarity 0.88)
- Harness the Memory: A Holistic Evaluation of Memory Substrates in Memory Agents (NeurIPS 2026 · similarity 0.87)
- Chain of Agents: Large Language Models Collaborating on Long-Context Tasks (NeurIPS 2024 · similarity 0.87)
- MEM1: Learning to Synergize Memory and Reasoning for Efficient Long-Horizon Agents (ICLR 2026 · similarity 0.86)
- M+: Extending MemoryLLM with Scalable Long-Term Memory (ICML 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
