Opening book details…
About this document
Refusal Falls Off A Cliff - How Safety Alignment Fails in Reasoning? by Massiel Méndez Jorge is a document available to read on EtoBox.
- Author
- Massiel Méndez Jorge
- Language
- EN
Opening book details…
Refusal Falls Off A Cliff - How Safety Alignment Fails in Reasoning? by Massiel Méndez Jorge is a document available to read on EtoBox.
English
This paper investigates the safety alignment failures in large reasoning models (LRMs) through the phenomenon termed