Агенты ИИ | AGI_and_RL

Replacing Rewards with Examples: Example-Based Policy Search via Recursive Classification https://arxiv.org/abs/2103.12656