Research map / Reinforcement learning / MDPs and planning
Sequential Transfer in Reinforcement Learning with a Generative Model
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 active information-seeking policy identification instead of jumpstart initialization | PTUM vs. MaxQInit · 12×12 two-room grid-world | episodes to find an optimal policy | Helped |
| Query informative state-action pairs rather than follow high-reward goals | PTUM vs. PEL and MaxQInit · Grid-world without a wall, multiple goals | convergence speed | Helped |
| Exploit estimated temporal task-transition correlations | Sequential transfer · Objectworld variant | normalized sample complexity | Helped |
| Use learned task models for transfer rather than oracle optimistic initialization | Sequential transfer vs. MaxQInit · Objectworld variant | near-optimal policy performance under the given sample budget | Helped |
Closest papers by meaning
- Breaking the Sample Size Barrier in Model-Based Reinforcement Learning with a Generative Model (NeurIPS 2020 · similarity 0.89)
- PAC Reinforcement Learning With an Imperfect Model (AAAI 2018 · similarity 0.89)
- Reinforcement Learning with Simple Sequence Priors (NeurIPS 2023 · similarity 0.88)
- Generative Modelling of Stochastic Actions with Arbitrary Constraints in Reinforcement Learning (NeurIPS 2023 · similarity 0.88)
- Prioritized Generative Replay (ICLR 2025 · similarity 0.88)
- Learning Generalizable and Composable Abstractions for Transfer in Reinforcement Learning (AAAI 2024 · similarity 0.87)
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.