Research map / Reinforcement learning / Visual and representation RL
Distributional Reward Decomposition for Reinforcement Learning
Zichuan Lin, Li Zhao, Derek Yang, Tao Qin, Tie-Yan Liu, Guangwen Yang
Read the paperCheck your idea against 137,217 papers
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 distributional reward decomposition with 2 or 3 channels to Rainbow | Rainbow · Atari games (six) | training performance | Improved |
| use 3 channels instead of 2 | DRDRL · Atari games | training performance | Mixed |
| use factorial approximation P(Zi|Zj)=P(Zi) instead of non-factorial full distribution (ablation) | DRDRL · Atari games | training performance | Improved |
| split state features into N pieces only (ablation) | DRDRL · Atari games | disentangled sub-returns | Hurt |
| add one-hot channel embedding multiplied with state features (ablation) | DRDRL · Atari games | training performance | Improved |
| add KL disentanglement term (ablation) | DRDRL · Atari games | training performance | Improved |
| use M atoms instead of M/N atoms per sub-distribution (ablation) | DRDRL · Atari games | training performance | Hurt |
Closest papers by meaning
- The Distributional Reward Critic Framework for Reinforcement Learning Under Perturbed Rewards (AAAI 2025 · similarity 0.89)
- RD$^2$: Reward Decomposition with Representation Decomposition (NeurIPS 2020 · similarity 0.89)
- Distributional Reinforcement Learning with Regularized Wasserstein Loss (NeurIPS 2024 · similarity 0.87)
- Intrinsic Benefits of Categorical Distributional Loss: Uncertainty-aware Regularized Exploration in Reinforcement Learning (NeurIPS 2025 · similarity 0.86)
- Information-Theoretic Reward Decomposition for Generalizable RLHF (NeurIPS 2025 · similarity 0.86)
- Decoding Global Preferences: Temporal and Cooperative Dependency Modeling in Multi-Agent Preference-Based Reinforcement Learning (AAAI 2024 · similarity 0.86)
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.
