169 citations · 375 across the 26 of their papers we have counts for
Showing cs.CRShow all
2 papers · 1 filter
cs.CR2026
Safety Anchor: Defending Harmful Fine-tuning via Geometric Bottlenecks
Guoxin Lu, Letian Sha, Qing Wang +4
The safety alignment of Large Language Models (LLMs) remains vulnerable to Harmful Fine-tuning (HFT). While existing defenses impose constraints on parameters, gradients, or intern…
cs.CR2025
Differentiated Directional Intervention A Framework for Evading LLM Safety Alignment
Peng Zhang, Peijie Sun
Safety alignment instills in Large Language Models (LLMs) a critical capacity to refuse malicious requests. Prior works have modeled this refusal mechanism as a single linear direc…