Research map / Reinforcement learning / Policy gradient and actor-critic
Performative Reinforcement 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 a less smooth environment response (for example, β=200) | Repeated Policy Optimization · Gridworld | convergence to a stable solution | Hurt |
| Increase regularization strength λ | Repeated Policy Optimization · Gridworld | convergence | Helped |
| Use a large gradient step size η | Repeated Gradient Ascent · Gridworld | convergence to a stable solution | Hurt |
| Use a step size η closer to 1/λ | Repeated Gradient Ascent · Gridworld | convergence speed | Helped |
| Increase the number of sampled trajectories m | Finite-sample repeated retraining · Gridworld | error around the stable solution | Helped |
| Solve the empirical Lagrangian directly instead of estimating the model and solving with estimated dynamics | Finite-sample repeated retraining · Gridworld | error around the stable solution | Helped |
Closest papers by meaning
- Performative Policy Gradient: Optimality in Performative Reinforcement Learning (ICML 2026 · similarity 0.94)
- On Corruption-Robustness in Performative Reinforcement Learning (AAAI 2025 · similarity 0.89)
- Performative Control for Linear Dynamical Systems (NeurIPS 2024 · similarity 0.88)
- Plug-in Performative Optimization (ICML 2024 · similarity 0.87)
- On the Convergence of Success Conditioning for Policy Optimization (NeurIPS 2026 · similarity 0.87)
- Reinforcement Learning with Random Time Horizons (ICML 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.