Notice: This website is continuously updated as new conference proceedings and journal volumes are published. Consequently, the figures presented here may differ from those reported in our publications.
A brief overview of our methodology is provided on the Methods page, along with references to publications documenting our approach and links to the code and data used in each study.
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.
Approximate Newton Methods and Their Local Convergence
Authors: Haishan Ye, Luo Luo, Zhihua Zhang
ICML 2017 | Venue PDF | LLM Run Details | Input Tokens: 17,568 Total number of tokens sent to the LLM as input for this paper's analysis. | Output Tokens: 5,277 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 | 7. Empirical Study In this section, we validate our theoretical results about sketched size of the sketch Newton, and sample size of regularized Newton, experimentally. Experiments for validating unnecessity of the Lipschitz continuity condition of 2F(x) are given in the supplementary materials. |
| Researcher Affiliation | Academia | 1Shanghai Jiao Tong University, Shanghai, China 2Peking University & Beijing Institute of Big Data Research, Beijing, China. |
| Pseudocode | Yes | Algorithm 1 Sketch Newton. ... Algorithm 2 Subsampled Newton. ... Algorithm 3 Regularized Subsample Newton. ... Algorithm 4 New Samp. |
| Open Source Code | No | The paper does not provide concrete access to source code for the methodology described. |
| Open Datasets | No | In our experiment, A is a 10000 54 matrix. We set the singular values σi of A as: σi = 1.2 i. ... Let A Rn d where n = 8000 and d = 5000. |
| Dataset Splits | No | The paper does not provide specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) for training, validation, or test sets, as it uses synthetically generated data for optimization problem instances rather than pre-existing datasets with standard splits. |
| Hardware Specification | No | The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific ancillary software details, such as library or solver names with version numbers, needed to replicate the experiment. |
| Experiment Setup | Yes | In our experiment, A is a 10000 54 matrix. We set the singular values σi of A as: σi = 1.2 i. Then the condition number of A is κ(A) = 1.254 = 1.8741 104. We use different sketch matrices in Sketch Newton (Algorithm 1) and set different values of the sketched size ℓ. ... Let A Rn d where n = 8000 and d = 5000. ... we set different sample sizes |S|. For each |S| we choose different regularizer terms α and different target ranks r. |