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 [1].
A Peer-review Look on Multi-modal Clustering: An Information Bottleneck Realization Method
Authors: Zhengzheng Lou, Hang Xue, Chaoyang Zhang, Shizhe Hu
ICML 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on eight multi-modal datasets suggest that PTIB can outperform the state-of-the-art multi-modal clustering methods. Experiments on eight multi-modal datasets demonstrate the superiority and effectiveness of the proposed method. |
| Researcher Affiliation | Academia | 1School of Computer Science and Artificial Intelligence, Zhengzhou University, Zhengzhou, China. |
| Pseudocode | Yes | Algorithm 1 The Proposed PTIB |
| Open Source Code | No | The paper does not contain an explicit statement about releasing source code or a link to a code repository. It does not mention supplementary materials for code either. |
| Open Datasets | Yes | Eight multi-modal datasets are used, including 20NG, COIL20, Event, Soccer, 17Flowers, 75Flowers, COIL100 and MMI. The brief information of them is summarized in Table 1 and their details are shown in the Appendix C.1. Footnotes provide URLs for all datasets, e.g., 'http://lig-membres.imag.fr/grimal/data.html' for 20NG dataset. |
| Dataset Splits | No | The paper mentions random partitioning of input 'X' into '|T|' data clusters at initialization, and conducting experiments 10 times, but does not specify train/test/validation splits or other specific data partitioning strategies for the datasets used. |
| Hardware Specification | Yes | All the compared methods and the proposed method are conducted in the same experimental environment, which is a desktop computer with Windows 10 operating system, 32GB RAM, and MATLAB 2021a. |
| Software Dependencies | Yes | All the compared methods and the proposed method are conducted in the same experimental environment, which is a desktop computer with Windows 10 operating system, 32GB RAM, and MATLAB 2021a. |
| Experiment Setup | Yes | There is only one parameter β in the proposed PTIB method, where we give the parameter settings as [10, 50, 100, 500, 700, 900, 1000]. Then we conduct extensive experiments on all the datasets with different settings to investigate the parameter sensitivity, and show the clustering performance with both Acc and NMI results in Figure 7. |