Research map / Bandits and online learning / Online convex optimization and games
Policy Optimization as Online Learning with Mediator Feedback
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 RANDOMIST instead of UCB1 | Policy optimization · LQG | cumulative regret | Helped |
| Set RANDOMIST perturbation scale to 1.1 instead of 8.1 (ablation) | RANDOMIST · LQG | cumulative regret | Helped |
| Use RANDOMIST instead of PGPE and PB-POIS | Continuous policy optimization · Mountain Car | cumulative return | Helped |
| Use RANDOMIST with perturbation scale 8.1 (ablation) | RANDOMIST · Illustrative 5-policy problems | cumulative regret | Helped |
| Use RANDOMIST with perturbation scale 1.1 instead of 8.1 (ablation) | RANDOMIST · Illustrative 5-policy problems | cumulative regret | Mixed |
| Use UCB1 instead of OPTIMIST and RANDOMIST when little information is shared among policies | Policy optimization · Illustrative 5-policy problems | cumulative regret | Helped |
| Use Gaussian-prior Thompson Sampling | Policy optimization · Illustrative 5-policy problems | cumulative regret | Helped |
| Use FTL under perfect mediator feedback | Finite policy space · Finite-policy PO | expected regret | Helped |
| Use OPTIMIST under perfect mediator feedback | Finite policy space · Finite-policy PO | expected regret | Helped |
| Use RANDOMIST under perfect mediator feedback | Finite policy space · Finite-policy PO | expected regret | Helped |
Closest papers by meaning
- Policy-Regret Minimization in Markov Games with Function Approximation (ICML 2025 · similarity 0.90)
- Rate-Optimal Policy Optimization for Linear Markov Decision Processes (ICML 2024 · similarity 0.88)
- Efficient Preference-Based Reinforcement Learning: Randomized Exploration meets Experimental Design (NeurIPS 2025 · similarity 0.88)
- Making RL with Preference-based Feedback Efficient via Randomization (ICLR 2024 · similarity 0.88)
- Near-optimal Regret Using Policy Optimization in Online MDPs with Aggregate Bandit Feedback (ICML 2025 · similarity 0.88)
- Reinforcement Learning with Random Time Horizons (ICML 2025 · 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.