Research map / Bandits and online learning / RL theory and MDPs
Linear function approximation in RL
Establish efficient optimistic algorithms for RL with linear and generalized linear features.
64 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2017–2026) · an approach in RL theory and MDPs, keyword cluster “rl · function approximation · linear function”
State History Tried Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
19 papers in 2023–24 and 17 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic held steady, from 26.0% to 28.3%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2017 · Unifying PAC and Regret: Uniform PAC Bounds for Episodic Reinforcement Learning (NeurIPS 2017 · 138 citations)
- 2020 · Learning Near Optimal Policies with Low Inherent Bellman Error (ICML 2020 · 61 citations)
- 2021 · Optimism in Reinforcement Learning with Generalized Linear Function Approximation (ICLR 2021 · 54 citations)
- 2022 · Provably Efficient Model-Free Constrained RL with Linear Function Approximation (NeurIPS 2022 · 6 citations)
- 2023 · Online Reinforcement Learning with Uncertain Episode Lengths (AAAI 2023 · 2 citations)
- 2024 · Demystifying Linear MDPs and Novel Dynamics Aggregation Framework (ICLR 2024 · 0 citations)
- 2025 · Logarithmic Regret for Online KL-Regularized Reinforcement Learning (ICML 2025 · 0 citations)
- 2026 · Exact Unlearning in Reinforcement Learning (ICML 2026 · 0 citations)
Earliest papers here
- Unifying PAC and Regret: Uniform PAC Bounds for Episodic Reinforcement Learning (NeurIPS 2017 · 138 citations)
- Learning Near Optimal Policies with Low Inherent Bellman Error (ICML 2020 · 61 citations)
- Learning with Good Feature Representations in Bandits and in RL with a Generative Model (ICML 2020 · 50 citations)
What has been tried
52 papers here report 219 measured changes (a component added, removed or swapped, compared with a baseline): helped 74%, no clear effect 8%, hurt 13%, mixed 5%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| Store only the previous episode instead of all historical data | Finite-horizon concurrent MDP · worst-case regret bound | a factor of \sqrt{K} | Hurt | Concurrent Reinforcement Learning with Aggregated States via Randomized Least Squares Value Iteration (ICML 2025 · 0 citations) |
| Do not use aggregation mappings | Block-RiverSwim · learning speed | Hurt | Demystifying Linear MDPs and Novel Dynamics Aggregation Framework (ICLR 2024 · 0 citations) |
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 36 papers; from OpenAlex and the papers' affiliations).
- Seoul National University · 4 papers
- Meta · 3 papers
- Google · 3 papers
- The Ohio State University · 2 papers
- University of California, Los Angeles · 2 papers
- Tsinghua University · 2 papers
- Cornell University · 2 papers
- UC Santa Barbara · 2 papers
Use it
Download BibTeX of 40 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Regret bounds with function approximation (79 papers)
- Stochastic shortest path (11 papers)
Elsewhere in the topic
- Average-reward and latent MDPs (57 papers)
- Non-stationary and structured MDPs (30 papers)
- Adversarial MDPs with bandit feedback (30 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).