‹ BackHN Continuity

Thread

"As a Language Model": Chat Template Switches LLM Self-Referential Voice

103 points · 110 comments · yu3zhou4

  1. cadamsdotcom · · focus · HN ↗
    Very cool innovation in steering - but a lot of introspection only emerges at the highest weight classes - this research would be fascinating to run on bigger models.
Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.