Research map / Reinforcement learning / MDPs and planning
On Reward-Free RL with Kernel and Neural Function Approximations: Single-Agent MDP and Markov Game
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 optimistic value iteration with kernel regression and an uncertainty bonus as the exploration reward | Single-agent MDP reward-free RL · Single-agent MDP, arbitrary extrinsic reward | sample complexity: \tilde{\mathcal{O}}(1/\varepsilon^2) | Helped |
| Use overparameterized neural regression with a neural uncertainty bonus as the exploration reward | Single-agent MDP reward-free RL · Single-agent MDP, arbitrary extrinsic reward | sample complexity: \tilde{\cO}(1/\varepsilon^2) | Helped |
| Extend optimistic reward-free exploration and planning to zero-sum games with kernel approximation | Zero-sum Markov game · Zero-sum Markov game, arbitrary extrinsic reward | sample complexity: \tilde{\cO}(1/\varepsilon^2) | Helped |
| Extend optimistic reward-free exploration and planning to zero-sum games with overparameterized neural approximation | Zero-sum Markov game · Zero-sum Markov game, arbitrary extrinsic reward | sample complexity: \tilde{\cO}(1/\varepsilon^2) | Helped |
| Use a finite-spectrum kernel with complexity parameter γ | Kernel reward-free RL · Finite-spectrum kernel; single-agent MDP and zero-sum Markov game | sample complexity: \tilde{\cO}(H^6 \gamma^3 /\varepsilon^2) | Helped |
| Use neural function approximation instead of exact kernel approximation | Reward-free RL · Single-agent MDP and zero-sum Markov game | suboptimality or Nash-equilibrium error bound: + H^2 \beta \iota | Mixed |
Closest papers by meaning
- Reward-Free Kernel-Based Reinforcement Learning (ICML 2024 · similarity 0.94)
- Reward-Free Exploration for Reinforcement Learning (ICML 2020 · similarity 0.93)
- On Reward-Free Reinforcement Learning with Linear Function Approximation (NeurIPS 2020 · similarity 0.93)
- Reward-Free RL is No Harder Than Reward-Aware RL in Linear Markov Decision Processes (ICML 2022 · similarity 0.92)
- Provably Efficient Reinforcement Learning with Kernel and Neural Function Approximations (NeurIPS 2020 · similarity 0.91)
- Exploration-Driven Policy Optimization in RLHF: Theoretical Insights on Efficient Data Utilization (ICML 2024 · similarity 0.91)
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.