Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The Shift Toward Open and Reproducible AI Research," Under Review, 2026. doi: 10.48550/arXiv.2606.16974..
Finite-Time Bounds for Average-Reward Fitted Q-Iteration
Authors: Jongmin Lee, Ernest K. Ryu
NeurIPS 2025 | Venue PDF | LLM Run Details | Input Tokens: 32,325 Total number of tokens sent to the LLM as input for this paper's analysis. | Output Tokens: 2,190 Total number of tokens produced by the LLM (including reasoning/thinking tokens) for this paper's analysis.
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | Our work does not include numerical experiments. |
| Researcher Affiliation | Academia | Jongmin Lee Seoul National University Department of Mathematical Sciences EMAIL Ernest K. Ryu UCLA Department of Mathematics EMAIL |
| Pseudocode | Yes | Algorithm 1 Anchored Fitted Q-Iteration (D, K, {Fi}K i=1{λi}K i=1) Input: D = {si, ai, ri, s i}n i=1, f0 = 0, K 1, {λi}K i=1 (0, 1) for k = 0, 1, . . . , K 1 do ˆTfk = argminf Fk+1 Pn i=1 f(si, ai) ri maxa A fk(s i, a) 2 fk+1 = (1 λk+1)f0 + λk+1 ˆTfk With f0 = 0, this is weight decay end for π(a | s) = argmaxa A f K(s, a) Output π, f K |
| Open Source Code | No | Our paper does not include experiments requiring code. |
| Open Datasets | No | Our work does not include numerical experiments. |
| Dataset Splits | No | Our work does not include numerical experiments. |
| Hardware Specification | No | Our paper does not include numerical experiments. |
| Software Dependencies | No | Our paper does not include numerical experiments. |
| Experiment Setup | No | Our paper does not include numerical experiments. |