Akademska digitalna zbirka SLovenije - logo

Search results

Basic search    Expert search   

Currently you are NOT authorised to access e-resources SI consortium. For full access, REGISTER.

1
hits: 1
1.
  • The Mysterious Case of Neuron 1512: Injectable Realignment Architectures Reveal Internal Characteristics of Meta's Llama 2 Model
    Smith, Brenden; Baker, Dallin; Chase, Clayton ... arXiv.org, 07/2024
    Paper, Journal Article
    Open access

    Large Language Models (LLMs) have an unrivaled and invaluable ability to "align" their output to a diverse range of human preferences, by mirroring them in the text they generate. The internal ...
Full text
Available for: NUK, UL, UM, UPUK

Load filters