reward maximization
A principle in reinforcement learning where an agent aims to maximize cumulative rewards based on its actions and experiences within an environment.
- CHPO: Constrained Hybrid-action Policy Optimization for Reinforcement Learning
- Modelling the control of offline processing with reinforcement learning
- Risk-aware Direct Preference Optimization under Nested Risk Measure
- Safe and Stable Control via Lyapunov-Guided Diffusion Models
- Structural Causal Bandits under Markov Equivalence
- When Can Model-Free Reinforcement Learning be Enough for Thinking?
- WorldModelBench: Judging Video Generation Models As World Models