Fetching the paper…

Efficient LLM Jailbreak via Adaptive Dense-to-sparse Constrained Optimization · Around