Research map / Learning theory and algorithms / Matrix and tensor methods
Hierarchical Subtask Discovery with Non-Negative Matrix Factorization
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 non-negative matrix factorization to approximate the desirability basis | MLMDP subtask discovery · Nested rooms, snake-rooms, and Taxi | qualitative subtask decomposition | Helped |
| Choose the task ensemble used to construct the desirability basis | MLMDP subtask discovery · Same domain with different task ensembles | subtask decomposition | Mixed |
| Iterate the decomposition across hierarchy levels | MLMDP subtask discovery · Nested rooms and 1D ring | hierarchical abstraction | Helped |
Closest papers by meaning
- Identifying Selections for Unsupervised Subtask Discovery (NeurIPS 2024 · similarity 0.90)
- Globally Optimal Hierarchical Reinforcement Learning for Linearly-Solvable Markov Decision Processes (AAAI 2022 · similarity 0.89)
- Conditional Diffusion Model for Multi-Agent Dynamic Task Decomposition (AAAI 2026 · similarity 0.88)
- Learning Task Belief Similarity with Latent Dynamics for Meta-Reinforcement Learning (ICLR 2025 · similarity 0.87)
- Prioritized Soft Q-Decomposition for Lexicographic Reinforcement Learning (ICLR 2024 · similarity 0.86)
- DISCOVER: Automated Curricula for Sparse-Reward Reinforcement Learning (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.