Special Topics in Deep Reinforcement Learning

Deep Reinforcement Learning

1. Value-based DRL Evolution

Core: Rainbow

Additional: DQN & Double DQN

2. Actor-Critic: Structure & Estimation

Core: A2C/A3C

Additional: GAE

3. Policy Optimization Stabilty

Core: TRPO & PPO

4. Continuous Control in DRL

Core: DDPG & SAC

AI Planning

FOL for goal-conditioned RL 

Sketch Decomposition via DRL

Model Aware Policy Transfer using Q-Learning

 

Exploration

NovelID

DEIR

Episodic Novelty Through Temporal Distance

Cell-Free Latent Go-Explore

 

Agentic RL

Group Relative Policy Optimization Algorithm

Tool-Integrated RL

Reflexion: Verbal Reinforcement Learning

Multi-Agent Reinforcement Learning