vix.ing · top · new · best · stats · spec

Yariv Barsheshat

  1. Innocuous-Seeming Data, Latent Ideology: Ideological Generalisation in Finetuned LLMs
    2026/07/16 by Robert Graham, Edward Stevinson, Yariv Barsheshat
    #cs.LG #cs.AI #cs.CL #cs.CY