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..
Understanding and Improving Feature Learning for Out-of-Distribution Generalization
Authors: Yongqiang Chen, Wei Huang, Kaiwen Zhou, Yatao Bian, Bo Han, James Cheng
NeurIPS 2023 | Venue PDF | LLM Run Details | Input Tokens: 55,976 Total number of tokens sent to the LLM as input for this paper's analysis. | Output Tokens: 4,905 Total number of tokens produced by the LLM (including reasoning/thinking tokens) for this paper's analysis.
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments show that Fe AT effectively learns richer features thus boosting the performance of various OOD objectives. |
| Researcher Affiliation | Collaboration | 1The Chinese University of Hong Kong 2RIKEN AIP EMAIL EMAIL Yatao Bian3, Bo Han4, James Cheng1 3Tencent AI Lab 4Hong Kong Baptist University EMAIL EMAIL |
| Pseudocode | Yes | Algorithm 1 Fe AT: Feature Augmented Training |
| Open Source Code | Yes | 1Code is available at https://github.com/LFhase/Fe AT. |
| Open Datasets | Yes | We conduct extensive experiments on both COLOREDMNIST [4, 16] and 6 datasets from the challenging benchmark, WILDS [39] |
| Dataset Splits | Yes | Table 8: A summary of datasets statistics from WILDS. Dataset # Examples # Domains train val test train val test |
| Hardware Specification | Yes | We run all the experiments on Linux servers with NVIDIA V100 graphics cards with CUDA 10.2. |
| Software Dependencies | Yes | We run all the experiments on Linux servers with NVIDIA V100 graphics cards with CUDA 10.2. |
| Experiment Setup | Yes | We use the Adam [37] optimizer with a learning rate of 1e 3 and a weight decay of 1e 3. |