value function
In reinforcement learning, the value function estimates the expected return or total future reward that an agent can achieve from a given state or action. It guides an agent's choices to maximize cumulative rewards.
- A Finite Sample Analysis of Distributional TD Learning with Linear Function Approximation
- A Temporal Difference Method for Stochastic Continuous Dynamics
- Certifying Stability of Reinforcement Learning Policies using Generalized Lyapunov Functions
- DAPO : Improving Multi-Step Reasoning Abilities of Large Language Models with Direct Advantage-Based Policy Optimization
- Deep learning for continuous-time stochastic control with jumps
- Option-aware Temporally Abstracted Value for Offline Goal-Conditioned Reinforcement Learning
- Preference Distillation via Value based Reinforcement Learning
- Retrosynthesis Planning via Worst-path Policy Optimisation in Tree-structured MDPs
- Scalable Policy-Based RL Algorithms for POMDPs
- Stop Summation: Min-Form Credit Assignment Is All Process Reward Model Needs for Reasoning
- Value Gradient Guidance for Flow Matching Alignment