Research map / Reinforcement learning / Offline RL
Exploit Reward Shifting in Value-Based Deep-RL: Optimistic Curiosity-Based Exploration and Conservative Exploitation via Linear Reward Shaping
Hao Sun, Lei Han, Rui Yang, Xiaoteng Ma, Jian Guo, Bolei Zhou
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add a positive reward shift | BCQ and CQL · offline RL (D4RL-style tasks incl. Hopper) | return | Improved |
| add a negative reward shift (ablation) | BCQ and CQL · offline RL | return | Hurt |
| add Random Reward Shift (3 Q-networks with shifts +0.5, 0, -0.5) | TD3 · MuJoCo continuous control (5 envs) | return | Improved |
| set all reward shifts to 0 (Ensemble TD3 vs RRS) (ablation) | RRS · MuJoCo continuous control | return | Improved |
| add a -0.5 reward shift | DQN · MountainCar and MiniGrid | exploration performance | Improved |
| use positive intrinsic reward (vanilla RND) with DQN | DQN · MountainCar and MiniGrid | exploration performance | Hurt |
| make RND intrinsic reward negative by subtracting a constant | RND with DQN · MountainCar and MiniGrid | exploration performance | Improved |
| increase negative shift magnitude from -1.0 to -1.5 (ablation) | RND with DQN · MountainCar and MiniGrid | exploration performance: -1.0 -> -1.5 | Improved |
Closest papers by meaning
- Random Latent Exploration for Deep Reinforcement Learning (ICML 2024 · similarity 0.88)
- Automatic Reward Shaping from Confounded Offline Data (ICML 2025 · similarity 0.88)
- Exploration from a Primal-Dual Lens: Value-Incentivized Actor-Critic Methods for Sample-Efficient Online RL (NeurIPS 2025 · similarity 0.87)
- MaxInfoRL: Boosting exploration in reinforcement learning through information gain maximization (ICLR 2025 · similarity 0.87)
- Hyper: Hyperparameter Robust Efficient Exploration in Reinforcement Learning (ICML 2025 · similarity 0.87)
- Beyond Optimism: Exploration With Partially Observable Rewards (NeurIPS 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
