LCZZZZ/model111

MemGen: Weaving Generative Latent Memory for Self-Evolving Agents

👋 Introduction

This repo is the official implementation of MemGen: Weaving Generative Latent Memory for Self-Evolving Agents

Inspired by the human brain’s ability to dynamically integrate memory and reasoning, MemGen introduces a novel framework that empowers AI agents to evolve through experience—without relying on rigid parameter updates or external databases.

Unlike traditional approaches, MemGen generates latent memory tokens directly within the model’s reasoning stream. It features:

• A Memory Trigger that decides when to recall memory.
• A Memory Weaver that synthesizes past experiences into compact, latent sequences—seamlessly enriching ongoing reasoning.

🌎 Setup

Create and activate the MemGen environment from the provided YAML file:

conda env create -f memgen.yaml
conda activate MemGen

🚀 Quick Start

🔧 Installation: Set Up Search Environment

Please follow the instructions in the Search-R1 to configure the retriever environment (optional).

---

▶️ How to Run

MemGen consists of two modules: Weaver and Trigger.
We follow a two-stage training approach, training each module separately.

Weaver Model

• Train the Weaver model

  bash weaver_train.sh
  

• Evaluate the Weaver model
  Before running, make sure to update LOAD_MODEL_PATH in eval.sh to point to the trained checkpoint: <weaver_dir>/model.safetensors

  bash eval.sh
  

Trigger Model

• Train the Trigger model

  bash trigger_train.sh
  

• Evaluate the Trigger model
  Before running, make sure to update LOAD_MODEL_PATH in eval.sh to point to the trained checkpoint: <trigger_dir>/model.safetensors

  bash eval.sh
  

🫡 Citation

If you find this repository helpful, a citation to our paper would be greatly appreciated:

@article{zhang2025memgen,
  title={MemGen: Weaving Generative Latent Memory for Self-Evolving Agents},
  author={Zhang, Guibin and Fu, Muxin and Yan, Shuicheng},
  journal={arXiv preprint arXiv:2509.24704},
  year={2025}
}

🙏 Acknowledgement

• We sincerely thank Search-R1 for open-sourcing their search web environment.
• We also extend our heartfelt thanks to LAVIS for their code framework design.