Research map / Reinforcement learning / Policy gradient and actor-critic
Reward Constrained Policy Optimization
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 adaptive RCPO rather than fixed-λ reward shaping | PPO on Swimmer-v2 · Swimmer-v2 | Reward: 11.7 -> 72.7 | Helped |
| Use adaptive RCPO rather than fixed-λ reward shaping | PPO on Walker2d-v2 · Walker2d-v2 | Reward: 823.5 -> 591.6 | Hurt |
| Use adaptive RCPO rather than fixed-λ reward shaping | PPO on Hopper-v2 · Hopper-v2 | Reward: 865.9 -> 1138.5 | Mixed |
| Use adaptive RCPO rather than fixed-λ reward shaping | PPO on Humanoid-v2 · Humanoid-v2 | Reward: 1151.8 -> 606.1 | Hurt |
| Use adaptive RCPO rather than fixed-λ reward shaping | PPO on HalfCheetah-v2 · HalfCheetah-v2 | Reward: -0.4 -> 1547.1 | Helped |
| Use adaptive RCPO rather than fixed-λ reward shaping | PPO on Ant-v2 · Ant-v2 | Reward: 1012.2 -> 1031.5 | Helped |
| Use the RCPO multi-timescale update instead of direct Lagrange-dual optimization | A2C on Mars Rover · Mars Rover, α = 0.01 and α = 0.5 | sample efficiency | Helped |
| Replace a fixed penalty coefficient with an adaptive coefficient | PPO on robotics domains · Mujoco robotics tasks | reward | Mixed |
Closest papers by meaning
- Feasible Policy Optimization for Safe Reinforcement Learning (NeurIPS 2026 · similarity 0.89)
- Constrained Reinforcement Learning Under Model Mismatch (ICML 2024 · similarity 0.89)
- Constrained Policy Optimization (ICML 2017 · similarity 0.89)
- Last-Iterate Global Convergence of Policy Gradients for Constrained Reinforcement Learning (NeurIPS 2024 · similarity 0.89)
- Reward Penalties on Augmented States for Solving Richly Constrained RL Effectively (AAAI 2024 · similarity 0.88)
- Going Beyond Heuristics by Imposing Policy Improvement as a Constraint (NeurIPS 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.