Robots That Know What to Ask: Recovering Misaligned Rewards through Targeted Explanations | Signal Canvas

Robots That Know What to Ask: Recovering Misaligned Rewards through Targeted Explanations | Signal Canvas | ScienceToStartup