Why Safeguarded Ships Run Aground? Aligned Large Language Models' Safety Mechanisms Tend to Be Anchored in The Template Region
Paper • 2502.13946 • Published • 10
Cooper Leong
cooperleong00
AI & ML interests
None yet
Organizations
None yet
authored 7
papers over 1 year ago
Instruct Once, Chat Consistently in Multiple Rounds: An Efficient Tuning Framework for Dialogue
Paper • 2402.06967 • Published
TokenSkip: Controllable Chain-of-Thought Compression in LLMs
Paper • 2502.12067 • Published • 4
Self-Detoxifying Language Models via Toxification Reversal
Paper • 2310.09573 • Published
E2CL: Exploration-based Error Correction Learning for Embodied Agents
Paper • 2409.03256 • Published • 1
Subtle Errors Matter: Preference Learning via Error-injected Self-editing
Paper • 2410.06638 • Published
Direct Preference Optimization Using Sparse Feature-Level Constraints
Paper • 2411.07618 • Published • 17