Discuss Alignment Tampering: How Reinforcement Learning from Human Feedback Is Exploited to Optimize Misaligned Biases
Share what worked, ask a question, or add context for other readers.
Sign in to join the discussion
Recent messages
No messages here yet.