Medical knowledge representation enhancement in large language models through clinical tokens optimization

Abstract During the training of medical large language models (LLMs), conventional tokenizers frequently segment domain-specific medical terms into multiple subword tokens, resulting in suboptimal recognition and representation of specialized vocabulary. As a consequence, the model encounters diffic...

Popoln opis

Bibliografske podrobnosti

Principais autores:	Qianqian Li, Jijun Tong, Shanna Liu, Chang Li, Jie Tang, Qingli Zhou
Format:	Artigo
Jezik:	Inglês
Izdano:	Nature Portfolio 2026-01-01
Serija:	Scientific Reports
Teme:	clinical tokens LLMs tokenizer natural language processing (NLP)
Online dostop:	https://doi.org/10.1038/s41598-026-37438-6
Oznake:	Označite Brez oznak, prvi označite!