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 embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
Hierarchical Shortest-Path Graph Kernel Network
Authors: Jiaxin Wang, Wenxuan Tu, Jieren Cheng
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experimental results demonstrate the effectiveness and superiority of the designed kernel and its corresponding learning framework compared to current competitors. Code is available at https://github.com/JXWANG-GRAPH/HSP-GKN. 4 Experiments In this section, we first evaluate the performance of the HSP kernel and its corresponding HSP-GKN framework on graph classification datasets. |
| Researcher Affiliation | Academia | Jiaxin Wang1 Wenxuan Tu2 Jieren Cheng1,2 1School of Cyberspace Security, Hainan University 2School of Computer Science and Technology, Hainan University EMAIL |
| Pseudocode | No | The paper describes its methodology using mathematical formulations and descriptive text, but it does not include any explicitly labeled 'Pseudocode' or 'Algorithm' blocks with structured, code-like steps. |
| Open Source Code | Yes | Code is available at https://github.com/JXWANG-GRAPH/HSP-GKN. |
| Open Datasets | Yes | We evaluate the proposed kernel and framework on 12 datasets from different domains and of varying scales, all included in the TUDatasets collection [43]. |
| Dataset Splits | Yes | We follow [50] and evaluate TUDatasets using a 10-fold cross-validation using their provided data splits for a fair comparison. For datasets without predefined splits, we adopt the splitting method provided in [50] and apply the same split across all baseline models. |
| Hardware Specification | Yes | All experiments were conducted on a system equipped with an 8-core Intel Xeon E5-2667v4@3.2GHz with 256 GB of RAM and an NVIDIA 3090 GPU. |
| Software Dependencies | No | The paper mentions specific software tools like 'Optuna [51]' and 'the library provided by [52]' (GraKel) but does not specify their version numbers, which are required for reproducible software dependency information. |
| Experiment Setup | Yes | For the HSP-GKN model configuration, we use a two-layer MLP, where each layer s dimension is half of the previous layer s dimension. For the selection of hyperparameters, we use Optuna [51] to tune the hyperparameters through an automated search process. |