From RL to IRL — Gaurav Mishra, Amazon AGI Lab
RL worked when the world was a game, and IRL starts when the game fights back. The talk catalogues what a reward function meets on contact with a real login screen.
Read articleTopic
All digests tagged Process Reward Modeling
Suggested route
Move from earlier context through the available material to the newest signals. This route is a reading aid, not a difficulty ranking.
RL worked when the world was a game, and IRL starts when the game fights back. The talk catalogues what a reward function meets on contact with a real login screen.
Read article