---
license: apache-2.0
language:
- en
base_model:
- Qwen/Qwen2.5-0.5B
---

## Model Description

This Memory Decoder model is trained on the Law domain and can be adapted to enhance any model in the Qwen2 and Qwen2.5 families. 

**Paper:** [Memory Decoder: A Pretrained, Plug-and-Play Memory for Large Language Models](https://www.arxiv.org/abs/2508.09874)

**GitHub:** [https://github.com/LUMIA-Group/MemoryDecoder](https://github.com/LUMIA-Group/MemoryDecoder/tree/main)

## Training & Evaluation Data

**Law Domain Dataset:** [AsyLex](https://huggingface.co/datasets/clairebarale/AsyLex)

**Test Split:** [MemoryDecoder-domain-data](https://huggingface.co/datasets/Clover-Hill/MemoryDecoder-domain-data)

## Performance Results

### Qwen2 Family

| Model | Base Model | Base + MemDec |
|-------|------------|---------------|
| Qwen2-0.5B | 10.23 | 4.57 |
| Qwen2-1.5B | 7.69 | 4.32 |
| Qwen2-7B | 5.92 | 4.00 |
| Qwen2-72B | 4.84 | 3.69 |

### Qwen2.5 Family

| Model | Base Model | Base + MemDec |
|-------|------------|---------------|
| Qwen2.5-0.5B | 9.86 | 4.57 |
| Qwen2.5-1.5B | 7.42 | 4.29 |
| Qwen2.5-3B | 6.68 | 4.16 |
| Qwen2.5-7B | 5.94 | 4.01 |
| Qwen2.5-14B | 5.35 | 3.86 |
| Qwen2.5-32B | 5.18 | 3.81 |
| Qwen2.5-72B | 4.84 | 3.70 |

*Perplexity scores on Law domain test set. Lower is better.*

## Citation

```bibtex
@article{cao2025memory,
  title={Memory decoder: A pretrained, plug-and-play memory for large language models},
  author={Cao, Jiaqi and Wang, Jiarui and Wei, Rubin and Guo, Qipeng and Chen, Kai and Zhou, Bowen and Lin, Zhouhan},
  journal={arXiv preprint arXiv:2508.09874},
  year={2025}
}
```

## Contact

For questions and support: maximus.cao@outlook.com