Research map / Reinforcement learning / Visual and representation RL
Large-Scale Retrieval for Reinforcement Learning
Peter Conway Humphreys, Arthur Guez, Olivier Tieleman, Laurent Sifre, Theophane Weber, Timothy P Lillicrap
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add retrieval of nearest-neighbor expert Go states (vs zeroed neighbors) | Retrieval-augmented MuZero-style agent · 9x9 Go test set | policy top-1 accuracy / value MSE | Improved |
| add retrieval to the agent | MCTS agent (200 sims) · Win rate vs Pachi | win rate | Improved |
| retrieve from only 1% of the dataset at evaluation (ablation) | Retrieval agent · 9x9 Go test set | prediction accuracy | Hurt |
| augment retrieval set with ~600k agent-Pachi game states at test time (ablation) | Retrieval agent with MCTS · Win rate vs Pachi | win rate | Improved |
| add neighbor randomisation during training (ablation) | Retrieval agent policy prior · Win rate vs Pachi | win rate | Improved |
| add neighbor dropout and neighbor regularisation (ablation) | Retrieval agent · Win rate vs Pachi | win rate | No clear effect |
| feed random neighbors at evaluation (vs relevant ones) (ablation) | Retrieval agent · Win rate vs Pachi | win rate | Hurt |
Closest papers by meaning
- REGENT: A Retrieval-Augmented Generalist Agent That Can Act In-Context in New Environments (ICLR 2025 · similarity 0.88)
- The Generalization Gap in Offline Reinforcement Learning (ICLR 2024 · similarity 0.87)
- DeepSearch: Overcome the Bottleneck of Reinforcement Learning with Verifiable Rewards via Monte Carlo Tree Search (ICLR 2026 · similarity 0.87)
- Search Self-Play: Pushing the Frontier of Agent Capability without Supervision (ICLR 2026 · similarity 0.87)
- XLand-100B: A Large-Scale Multi-Task Dataset for In-Context Reinforcement Learning (ICLR 2025 · similarity 0.86)
- Horizon Reduction Makes RL Scalable (NeurIPS 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.
