Research map / Reinforcement learning / Policy gradient and actor-critic
V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control
H. Francis Song, Abbas Abdolmaleki, Jost Tobias Springenberg, Aidan Clark, Hubert Soyer, Jack W. Rae, Seb Noury, Arun Ahuja, Siqi Liu, Dhruva Tirumala, Nicolas Heess, Dan Belov and 2 more
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 |
|---|---|---|---|
| Use V-MPO instead of IMPALA | Atari-57 multi-task · Atari-57 | human-normalized median: 117.56 -> 155.70 | Improved |
| Use the top 50% of batch advantages for the E-step (ablation) | V-MPO · policy improvement | learning performance | Improved |
| Remove the parametric KL constraint (ablation) | V-MPO on humanoid gaps from pixels · Humanoid gaps | learning performance | Hurt |
| Use V-MPO rather than previously reported methods | 22-dimensional humanoid from full-state observations · Humanoid run | asymptotic return | Improved |
| Use V-MPO rather than previously reported methods | OpenAI Gym tasks · OpenAI Gym | asymptotic performance | Improved |
Closest papers by meaning
- Maximum a Posteriori Policy Optimisation (ICLR 2018 · similarity 0.92)
- TOPPO: Rethinking PPO for Multi-Task Reinforcement Learning with Critic Balancing (NeurIPS 2026 · similarity 0.87)
- Diffusion Policy Policy Optimization (ICLR 2025 · similarity 0.87)
- GenPO: Generative Diffusion Models Meet On-Policy Reinforcement Learning (NeurIPS 2025 · similarity 0.87)
- Efficient Test-Time Adaptation For Robot Policies (NeurIPS 2026 · similarity 0.86)
- Accelerating Visual-Policy Learning through Parallel Differentiable Simulation (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.
