AI safety and alignment · MLD-1323
Weak-to-strong generalization
A stronger learner improving beyond the capabilities or reliability of its weaker supervisor.
Original reference page · detailed expansion pending. Original definitions, examples, notes and citations are preserved below. New examples are illustrative, not benchmark results.
Original example
Train from imperfect supervision and evaluate independent capability.
Original caveat
Successful task generalization does not establish safe alignment in general.
Context and aliases
Level: Advanced. Stages: Training · Evaluation · Research · Governance.
Tags: Safety · RL · SLM · LLM.
Aliases: None recorded.