Research map / Reinforcement learning / Visual and representation RL
Continual Auxiliary Task Learning
Matthew K McLeod, Chunlok Lo, Matthew Kyle Schlegel, Andrew Jacobsen, Raksha Kumaraswamy, Martha White, Adam M White
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 |
|---|---|---|---|
| Use SF-NR instead of Tree Backup for GVF learning | Continual Auxiliary Task Learning · Tabular TMaze | prediction RMSE | Improved |
| Use SF-NR instead of Tree Backup for GVF learning | Continual Auxiliary Task Learning · Continuous TMaze | prediction RMSE | Improved |
| Use GPI instead of Sarsa for behavior control | Continual Auxiliary Task Learning · Tabular TMaze | prediction RMSE | No clear effect |
| Use GPI instead of Sarsa for behavior control | Continual Auxiliary Task Learning · Continuous TMaze | prediction RMSE | Improved |
| Use GPI instead of a random behavior policy | Continual Auxiliary Task Learning · Mountain Car | GVF prediction RMSE | Improved |
| Use replay for the stationary successor-feature component (ablation) | SF-NR · Continuous TMaze | prediction RMSE | Improved |
| Add replay to Tree Backup learning (ablation) | Tree Backup · Continuous TMaze | prediction RMSE | Hurt |
| Add interest functions to focus predictions on relevant states (ablation) | GVF learners · Open 2D World | prediction RMSE | Improved |
| Use emphatic reweighting with interest instead of interest alone (ablation) | GVF learners · Open 2D World | prediction RMSE | Mixed |
| Use prior corrections to reweight learning toward the target policy (ablation) | GVF learners · Off-policy learning with exploratory behavior | learning effectiveness | Hurt |
| Use reward features that generalize excessively across states (ablation) | SF-NR · Continuous TMaze | prediction RMSE | Hurt |
| Learn values directly with LSTD instead of separating successor features and cumulant learning | GVF learners · Tabular TMaze | prediction RMSE | Hurt |
Closest papers by meaning
- Adaptive Auxiliary Task Weighting for Reinforcement Learning (NeurIPS 2019 · similarity 0.90)
- Reinforcement Learning via Auxillary Task Distillation (ECCV 2024 · similarity 0.89)
- Learning by Playing Solving Sparse Reward Tasks from Scratch (ICML 2018 · similarity 0.87)
- Learning Novel Policies For Tasks (ICML 2019 · similarity 0.87)
- Hierarchical Reinforcement Learning with Advantage-Based Auxiliary Rewards (NeurIPS 2019 · similarity 0.87)
- Boosting Reinforcement Learning with Strongly Delayed Feedback Through Auxiliary Short Delays (ICML 2024 · similarity 0.85)
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.
