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.
Local Differential Privacy-Preserving Spectral Clustering for General Graphs
Authors: Sayan Mukherjee, Vorapong Suppakitpaisarn
TMLR 2025 | Venue PDF | LLM Run Details | Input Tokens: 19,574 Total number of tokens sent to the LLM as input for this paper's analysis. | Output Tokens: 2,530 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 | Empirical tests further corroborate these theoretical findings. |
| Researcher Affiliation | Academia | Sayan Mukherjee EMAIL The University of Tokyo Vorapong Suppakitpaisarn EMAIL The University of Tokyo |
| Pseudocode | Yes | Compute the second smallest eigenvector v = [v1, . . . , vn] of LG using the Lanczos algorithm or an approximation method, such as the one proposed in Adil & Saranurak (2024). Let v1, . . . , vn be distinct nodes of V such that vv1 vvn. Return the cut (S, S), where S = {v1, . . . , i0} and i0 = arg min 1 i n ÎąG(v1, . . . , vi). |
| Open Source Code | No | The paper does not contain any explicit statements or links indicating that open-source code for the described methodology is provided. |
| Open Datasets | Yes | We conduct experiments on real social networks to verify our theoretical results. In this work, we mainly use the network called Social circles: Facebook obtained from the Stanford network analysis project (SNAP), detailed in Leskovec & Mcauley (2012). |
| Dataset Splits | Yes | For each probability p and graph, we create 100 random graphs F with the given probability. |
| Hardware Specification | No | The paper does not contain any specific hardware details (e.g., GPU/CPU models, processor types, memory amounts) used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific ancillary software details (e.g., library or solver names with version numbers) needed to replicate the experiment. |
| Experiment Setup | Yes | We examine p {0.0001q : 1 q 50}. For each probability p and graph, we create 100 random graphs F with the given probability. Note that the original graph is represented by G. We then compute the difference between the clustering results of G (represented by SC2(G)) and that of G F (represented by SC2(G F)). |