ashtianicode / gradient_hacking

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

gradient_hacking

Does the llm preserve its goal under the pressure of optimization (fine-tuning)

About


Languages

Language:Python 100.0%