The Embedding Report.

Entities · Labs

HarmBench

2 articles tagged with this entity.

Breaking Safety at the Token Boundary: How BPE Tokenization Creates Exploitable Gaps in LLM Alignment

via export.arxiv.org · Global · 21d ago
Beyond Pass/Fail: Using Process Mining to Understand How LLMs Resist (and Fail) Red Team Attacks

via export.arxiv.org · Global · 45d ago