Alignment Tampering: How Reinforcement Learning from Human Feedback Is Exploited to Optimize Misaligned Biases
Paper • 2605.27355 • Published • 7
Dongyoon Hahm
Hahmdong
AI & ML interests
AI Safety
Recent Activity
updated a model about 6 hours ago
Hahmdong/PERSONA-qwen3-4b_quirky published a model about 6 hours ago
Hahmdong/PERSONA-qwen3-4b_quirky updated a model 6 days ago
Hahmdong/RMOOD-qwen3-4b-it-skywork-bias-analysis-emoji-p10-p100Organizations
None yet