Research map / Reinforcement learning / MDPs and planning
Provably Efficient Exploration for Reinforcement Learning Using Unsupervised 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 URL with unsupervised decoding and OracleQ for tabular RL | LockBernoulli · LockBernoulli | average reward: 80% | Helped |
| Use URL with unsupervised decoding and OracleQ for tabular RL | LockGaussian · LockGaussian, H=20, α=0.2 and 0.5 | average reward: >75% | Helped |
| Use URL instead of PCID | LockBernoulli and LockGaussian · LockBernoulli and LockGaussian | average reward | Mixed |
| Use URL instead of QLearning-lat | LockBernoulli · LockBernoulli, especially H=20 | average reward | Helped |
| Use URL instead of observation-only OracleQ or QLearning | LockBernoulli · LockBernoulli | average reward | Helped |
| Apply URL in LockGaussian despite violated BMDP assumptions | URL · LockGaussian | average reward: >75% | Helped |
Closest papers by meaning
- Model-Free Active Exploration in Reinforcement Learning (NeurIPS 2023 · similarity 0.90)
- Exploration in Structured Reinforcement Learning (NeurIPS 2018 · similarity 0.90)
- Task-agnostic Exploration in Reinforcement Learning (NeurIPS 2020 · similarity 0.90)
- Reward-Free Exploration for Reinforcement Learning (ICML 2020 · similarity 0.90)
- Provably efficient RL with Rich Observations via Latent State Decoding (ICML 2019 · similarity 0.90)
- Beyond Optimism: Exploration With Partially Observable Rewards (NeurIPS 2024 · similarity 0.89)
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.