README.md

---
license: mit
datasets:
- chengpingan/PIP-KAG
language:
- en
base_model:
- meta-llama/Meta-Llama-3-8B
library_name: transformers
---

# 🤖 PIP-KAG: Mitigating Knowledge Conflicts in Knowledge-Augmented Generation via Parametric Pruning

This is the official model for **[PIP-KAG: Mitigating Knowledge Conflicts in Knowledge-Augmented Generation via Parametric Pruning](https://arxiv.org/pdf/2502.15543)**.  

The PIP-KAG model is designed to address **knowledge conflicts** in **knowledge-augmented generation** tasks by leveraging a **parametric pruning** strategy, improving the **contextual faithfulness** of language models during knowledge-intensive generation.

---

## 📚 **Paper**
For a detailed explanation of the methodology and experiments, please refer to our paper:  
[**PIP-KAG: Mitigating Knowledge Conflicts in Knowledge-Augmented Generation via Parametric Pruning**](https://arxiv.org/abs/2502.15543)

---

## 📊 Reproduce the Results
To reproduce the experiments and benchmarks from the paper, follow the instructions provided in the official GitHub repository:
[👉 GitHub: OpenBMB/PIP-KAG](https://github.com/OpenBMB/PIP-KAG).

## 📁 Model Details
- Model Name: PIP-KAG-7B
- Architecture: LLaMA3-8B-Instruct with Parametric Pruning
- Training Data: [CoConflictQA](https://huggingface.co/datasets/chengpingan/PIP-KAG) Dataset
- Pretrained Tasks: Knowledge-Augmented Generation, Contextual Faithfulness Evaluation

## 🔖 Citation
If you use PIP-KAG in your work, please consider citing our paper:
```
@misc{huang2025pipkagmitigatingknowledgeconflicts,
      title={PIP-KAG: Mitigating Knowledge Conflicts in Knowledge-Augmented Generation via Parametric Pruning}, 
      author={Pengcheng Huang and Zhenghao Liu and Yukun Yan and Xiaoyuan Yi and Hao Chen and Zhiyuan Liu and Maosong Sun and Tong Xiao and Ge Yu and Chenyan Xiong},
      year={2025},
      eprint={2502.15543},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2502.15543}, 
}

```