Research map / Reinforcement learning / Policy gradient and actor-critic
Direct Policy Gradients: Direct Optimization of Policies in Discrete Action Spaces
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use DirPG instead of full-bandit UCB | Combinatorial bandits · Combinatorial bandits | average return | Helped |
| Use DirPG instead of semi-bandit UCB | Combinatorial bandits · Combinatorial bandits | average return | Mixed |
| Set the search priority weight to α=0.3 rather than α=0 (ablation) | DirPG · MiniGrid-MultiRoom-N6-v0 | average return | Helped |
| Increase the search priority weight beyond α=0.3 (ablation) | DirPG · MiniGrid-MultiRoom-N6-v0 | average return | Hurt |
| Use α=1 (A* search) instead of a lower search-priority weight (ablation) | DirPG · MiniGrid-MultiRoom-N6-v0 | completed trajectories | Hurt |
| Use positive ε to favor risk-seeking behavior (ablation) | DirPG · DeepSea | probability assigned to RRRR | Helped |
| Use negative ε to favor risk avoidance (ablation) | DirPG · DeepSea | probability assigned to LLLL | Helped |
| Include the control-variate term (ablation) | DirPG · MiniGrid | update variance | Helped |
Closest papers by meaning
- PODS: Policy Optimization via Differentiable Simulation (ICML 2021 · similarity 0.89)
- Low-Switching Policy Gradient with Exploration via Online Sensitivity Sampling (ICML 2023 · similarity 0.89)
- Ranking Policy Gradient (ICLR 2020 · similarity 0.89)
- Deterministic Policy Optimization by Combining Pathwise and Score Function Estimators for Discrete Action Spaces (AAAI 2018 · similarity 0.89)
- Diffusion Policy Policy Optimization (ICLR 2025 · similarity 0.88)
- DGPO: Discovering Multiple Strategies with Diversity-Guided Policy Optimization (AAAI 2024 · similarity 0.88)
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.