Research map / Reinforcement learning / Policy gradient and actor-critic
Trust the Model When It Is Confident: Masked Model-based Actor-Critic
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 masked model rollouts instead of unmasked rollouts | M2AC · MuJoCo continuous control | return with long rollouts | Helped |
| masking rate w close to 0.9 in long rollouts (ablation) | M2AC · HalfCheetah-Noisy2 | return | Hurt |
| penalty coefficient alpha 0.001 vs 0.01 or 0 (ablation) | M2AC · HalfCheetah-Noisy2 | return | Helped |
| non-stop rollout mode instead of hard-stop (ablation) | M2AC · HalfCheetah-Noisy2 | return | Helped |
| One-vs-Rest uncertainty instead of One-vs-All or negative likelihood (ablation) | M2AC · HalfCheetah-Noisy2 | return | Helped |
| M2AC vs MBPO with H_max=1 | M2AC · MuJoCo continuous control | return | No clear effect |
| M2AC vs MBPO in noisy Walker2d | M2AC · Walker2d noisy | return | Helped |
Closest papers by meaning
- Trust the Model Where It Trusts Itself - Model-Based Actor-Critic with Uncertainty-Aware Rollout Adaption (ICML 2024 · similarity 0.84)
- Double Check Your State Before Trusting It: Confidence-Aware Bidirectional Offline Model-Based Imagination (NeurIPS 2022 · similarity 0.83)
- Towards General-Purpose Model-Free Reinforcement Learning (ICLR 2025 · similarity 0.80)
- Actor-Critics Can Achieve Optimal Sample Efficiency (ICML 2025 · similarity 0.80)
- Offline-Boosted Actor-Critic: Adaptively Blending Optimal Historical Behaviors in Deep Off-Policy RL (ICML 2024 · similarity 0.80)
- Towards Robust Model-Based Reinforcement Learning Against Adversarial Corruption (ICML 2024 · similarity 0.80)
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.