---
license: apache-2.0
---

## Dataset Description

This dataset is a preprocessed version of the English Wikipedia snapshot from December 2021, processed using the `preprocess_dataset.py` script provided in the repository below.

- **Paper:** [MLP Memory: A Retriever-Pretrained Memory for Large Language Models](https://arxiv.org/pdf/2508.01832)
  
- **GitHub:** [https://github.com/Rubin-Wei/MLPMemory](https://github.com/Rubin-Wei/MLPMemory)

- **Dataset Source:** [English Wikipedia (December 2021)](https://github.com/facebookresearch/atlas?tab=readme-ov-file#corpora)

- **Tokenizer:** Mistral-7B-v0.3

Two key preprocessing parameters used are:

- `block_size`: 2048  
- `stride`: 1024

## Data Splits

The dataset is organized into `train` and `test` splits as follows:
```
├── test
├── train
├── dataset_dict.json
└── dstore_summary.json
```

The train/test split was created using the following code:
```python
from datasets import load_dataset

dataset = load_dataset('json', data_files=dataset_path)
split_dataset = dataset['train'].train_test_split(
    test_size=0.002,
    seed=42,
    shuffle=True
)
train_dataset = split_dataset['train']
test_dataset = split_dataset['test']
raw_datasets = DatasetDict({"train": train_dataset, "test": test_dataset})
```


## Citation

```bibtex
@inproceedings{Wei2025MLPMA,
  title={MLP Memory: A Retriever-Pretrained Memory for Large Language Models},
  author={Rubin Wei and Jiaqi Cao and Jiarui Wang and Jushi Kai and Qipeng Guo and Bowen Zhou and Zhouhan Lin},
  year={2025},
  url={https://api.semanticscholar.org/CorpusID:281658735}
}
```

## Contact

For questions and discussions, feel free to email: **weirubinn@gmail.com**