English

Chunk-Distilled Language Modeling

Computation and Language 2025-01-03 v1 Artificial Intelligence

Abstract

We introduce Chunk-Distilled Language Modeling (CD-LM), an approach to text generation that addresses two challenges in current large language models (LLMs): the inefficiency of token-level generation, and the difficulty of adapting to new data and knowledge. Our method combines deep network-based LLMs with a straightforward retrieval module, which allows the generation of multi-token text chunks at a single decoding step. Our retrieval framework enables flexible construction of model- or domain-specific datastores, either leveraging the internal knowledge of existing models, or incorporating expert insights from human-annotated corpora. This adaptability allows for enhanced control over the language model's distribution without necessitating additional training. We present the CD-LM formulation along with performance metrics demonstrating its ability to improve language model performance and efficiency across a diverse set of downstream tasks. Code and data will be made publicly available.

Keywords

Cite

@article{arxiv.2501.00343,
  title  = {Chunk-Distilled Language Modeling},
  author = {Yanhong Li and Karen Livescu and Jiawei Zhou},
  journal= {arXiv preprint arXiv:2501.00343},
  year   = {2025}
}
R2 v1 2026-06-28T20:53:12.198Z