Toward Open Weight Models Without Risks: Separating Public and Private Capabilities in LLMs
Paper LLMs & Texto

Toward Open Weight Models Without Risks: Separating Public and Private Capabilities in LLMs

Tiered Language Models (TLMs) provide a framework for releasing large language models with configurable capability levels through secret keys that modify computation graphs while m…

Hugging Face · Daily Papers ·Charbel El Feghali, Arkil Patel · ·▲ 1 upvotes

Este artigo está em destaque na seleção diária de papers do Hugging Face, curada pela comunidade de pesquisa em IA.

Autores: Charbel El Feghali, Arkil Patel, Nicholas Meade, Spandana Gella, Verna Dankers, Siva Reddy

  • 1 upvotes da comunidade
  • Temas: Tiered Language Models, secret key, parameter subset, computation graph, pretraining, fine-tuning

Resumo

Resumo original (em inglês), extraído do paper:

Tiered Language Models (TLMs) provide a framework for releasing large language models with configurable capability levels through secret keys that modify computation graphs while maintaining public model integrity.

Ler o paper completo no Hugging Face →

compartilhar: