Weiyifan/Llama3-8B-senator

Structural Entropy Guided Agent for Detecting and Repairing Knowledge Deficiencies in LLMs

🔔 Code • 📃 Paper • 🤗 Dataset

This is a model of llama3-8B trained based on the senator framework. You can also find more detailed descriptions at https://github.com/weiyifan1023/senator/edit/main/README.md.

Citation

If you find our paper inspiring and have utilized it in your work, please cite our paper.

@inproceedings{
wei2025structural,
title={Structural Entropy Guided Agent for Detecting and Repairing Knowledge Deficiencies in {LLM}s},
author={Yifan Wei and Xiaoyan Yu and Tengfei Pan and Angsheng Li and Li Du},
booktitle={The Thirty-ninth Annual Conference on Neural Information Processing Systems},
year={2025},
url={https://openreview.net/forum?id=hTGqC1h8Ig}
}

Acknowledgements

Thanks to the authors of KG-RAG and DAMe for releasing their code for retrieving the SPOKE KG and evaluating SE on the graph. Much of this codebase has been adapted from their codes.

Contact

weiyifan@buaa.edu.cn