LLMs Encode Harmfulness and Refusal Separately
Paper
•
2507.11878
•
Published
•
1
jc z
nickzzZzz
·
AI & ML interests
None yet
Recent Activity
upvoted
a
paper
about 2 months ago
Can Aha Moments Be Fake? Identifying True and Decorative Thinking Steps
in Chain-of-Thought
updated
a dataset
2 months ago
nickzzZzz/Identify_true_decorative_thinking
published
a dataset
3 months ago
nickzzZzz/Identify_true_decorative_thinking
Organizations
authored
a
paper
5 months ago