Offline Multitask Representation Learning for Reinforcement Learning
Authors: Haque Ishfaq, Thanh Nguyen-Tang, Songtao Feng, Raman Arora, Mengdi Wang, Ming Yin, Doina Precup
NeurIPS 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | We theoretically investigate offline multitask low-rank RL, and propose a new algorithm called MORL for offline multitask representation learning. Furthermore, we examine downstream RL in reward-free, offline and online scenarios, where a new task is introduced to the agent that shares the same representation as the upstream offline tasks. Our theoretical results demonstrate the benefits of using the learned representation from the upstream offline task instead of directly learning the representation of the low-rank model. |
| Researcher Affiliation | Academia | Haque Ishfaq Mila, Mc Gill University haque.ishfaq@mail.mcgill.ca Thanh Nguyen-Tang Johns Hopkins University nguyent@cs.jhu.edu Songtao Feng University of Florida sfeng1@ufl.edu Raman Arora Johns Hopkins University arora@cs.jhu.edu Mengdi Wang Princeton University mengdiw@princeton.edu Ming Yin Princeton University my0049@princeton.edu Doina Precup Mila, Mc Gill University dprecup@cs.mcgill.ca |
| Pseudocode | Yes | Algorithm 1 Multitask Offline Representation Learning (MORL) ... Algorithm 2 Downstream Reward-Free Exploration: Exploration Phase ... Algorithm 3 Downstream Reward-Free Exploration: Planning Phase |
| Open Source Code | No | The paper does not contain any explicit statements or links indicating the release of source code for the described methodology. It is a theoretical paper. |
| Open Datasets | No | This is a theoretical paper that does not involve experimental evaluation with datasets, and thus does not mention publicly available training datasets. |
| Dataset Splits | No | This is a theoretical paper and does not involve experimental evaluation or dataset splits for training, validation, or testing. |
| Hardware Specification | No | This is a theoretical paper and does not describe any experimental setup that would require hardware specifications. |
| Software Dependencies | No | This is a theoretical paper and does not describe any experimental setup that would require software dependencies with specific versions. |
| Experiment Setup | No | This is a theoretical paper and focuses on algorithm design and theoretical analysis, without describing any experimental setup details such as hyperparameters or training configurations. |