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..
Causal Spatio-Temporal Prediction: An Effective and Efficient Multi-Modal Approach
Authors: Yuting Huang, Ziquan Fang, Zhihao Zeng, Lu Chen, Yunjun Gao
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on 4 real-world datasets show that E2-CSTP significantly outperforms 9 state-of-the-art methods, achieving up to 9.66% improvements in accuracy as well as 17.37% 56.11% reductions in computational overhead. |
| Researcher Affiliation | Academia | Yuting Huang, Ziquan Fang , Zhihao Zeng, Lu Chen, Yunjun Gao Zhejiang University EMAIL |
| Pseudocode | Yes | Algorithm 1: Training procedure of E2-CSTP |
| Open Source Code | Yes | We provide the source code and datasets at https://github.com/ ZJU-DAILY/E2-CSTP. |
| Open Datasets | Yes | We collect 4 datasets to evaluate the proposed E2-CSTP framework: Terra [4], Bj TT [59], Green Earth Net [3], and Bike NYC [60]. |
| Dataset Splits | Yes | The datasets are chronologically divided into training, validation, and test sets in an 8:1:1 ratio. |
| Hardware Specification | Yes | All experiments are conducted on a Rocky Linux 8.8 server equipped with NVIDIA A40 GPUs. |
| Software Dependencies | Yes | We implement E2-CSTP using Python 3.8.20 and Py Torch 2.0.1. |
| Experiment Setup | Yes | Model training is performed using the Adam optimizer with an initial learning rate of 0.001, which decays by a factor of 0.5 every 5 epochs. We adopt early stopping based on the validation loss with a patience of 10 epochs to prevent overfitting and ensure stable convergence. |