Research map / Reinforcement learning / Policy gradient and actor-critic
On the Role of Weight Sharing During Deep Option Learning
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 OCPG rather than OC | Option-critic on Alien · Alien 10M | evaluation reward: 491.5 -> 663.0 | Helped |
| Use OCPG rather than OC | Option-critic on Alien · Alien 25M | evaluation reward: 618.8 -> 791.4 | Helped |
| Use OCPG rather than OC | Option-critic on Amidar · Amidar 10M | evaluation reward: 45.1 -> 219.3 | Helped |
| Use OCPG rather than OC | Option-critic on Amidar · Amidar 25M | evaluation reward: 89.4 -> 141.9 | Helped |
| Use OCPG rather than OC | Option-critic on Asterix · Asterix 10M | evaluation reward: 409.7 -> 1213.4 | Helped |
| Use OCPG rather than OC | Option-critic on Asterix · Asterix 25M | evaluation reward: 1042.0 -> 1679.8 | Helped |
| Use OCPG rather than OC | Option-critic on Boxing · Boxing 10M | evaluation reward: -37.7 -> -41.4 | Hurt |
| Use OCPG rather than OC | Option-critic on Boxing · Boxing 25M | evaluation reward: 32.9 -> 90.8 | Helped |
| Use OCPG rather than OC | Option-critic on Jamesbond · Jamesbond 10M | evaluation reward: 7.1 -> 118.6 | Helped |
| Use OCPG rather than OC | Option-critic on Jamesbond · Jamesbond 25M | evaluation reward: 29.3 -> 317.7 | Helped |
| Use OCPG rather than OC | Option-critic on Tennis · Tennis 10M | evaluation reward: -17.9 -> -14.4 | Helped |
| Use OCPG rather than OC | Option-critic on Tennis · Tennis 25M | evaluation reward: -19.9 -> -13.1 | Helped |
Closest papers by meaning
- The Option-Critic Architecture (AAAI 2017 · similarity 0.90)
- Flexible Option Learning (NeurIPS 2021 · similarity 0.90)
- Natural Option Critic (AAAI 2019 · similarity 0.88)
- Reward-Respecting Subtasks for Model-Based Reinforcement Learning (Abstract Reprint) (AAAI 2024 · similarity 0.87)
- Adversarial Policy Optimization for Offline Preference-based Reinforcement Learning (ICLR 2025 · similarity 0.86)
- Option-aware Temporally Abstracted Value for Offline Goal-Conditioned Reinforcement Learning (NeurIPS 2025 · 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.