Learning to Remember: End-to-End Training of Memory Agents for Long-Context Reasoning
cs.LG, cs.AI
Submitted: 2026-02-13
Updated: 2026-09-01
Comments: Accepted at EMNLP 2026 Main Conference. Camera-ready version
Code: https://github.com/ictnlp/unified-memory-agent
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Evaluating Memory in LLM Agents via Incremental Multi-Turn Interactions
- DeepSeek-V3 Technical Report
- Convomem Benchmark: Why Your First 150 Conversations Don't Need RAG
- Memory Transformer
- Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory
- Qwen3-VL Technical Report
- PerLTQA: A Personal Long-Term Memory Dataset for Memory Classification, Retrieval, and Synthesis in Question Answering
- In-context Autoencoder for Context Compression in a Large Language Model
- Zep: A Temporal Knowledge Graph Architecture for Agent Memory
- HybridFlow: A Flexible and Efficient RLHF Framework
- MemGen: Weaving Generative Latent Memory for Self-Evolving Agents
- MEM1: Learning to Synergize Memory and Reasoning for Efficient Long-Horizon Agents
- MIRIX: Multi-Agent Memory System for LLM-Based Agents
- Retrieval meets Long Context Large Language Models
- Memory-R1: Enhancing Large Language Model Agents to Manage and Utilize Memories via Reinforcement Learning
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks