Physics-Integrated Variational Autoencoders for Robust and Interpretable Generative Modeling

Authors: Naoya Takeishi, Alexandros Kalousis

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We performed experiments on two synthetic datasets and two real-world datasets, for which we prepared instances of physics-integrated VAEs.
Researcher Affiliation Academia University of Applied Sciences and Arts Western Switzerland (HES-SO) Geneva, Switzerland
Pseudocode No The paper does not contain any pseudocode or clearly labeled algorithm blocks.
Open Source Code No The paper does not provide concrete access to its source code, nor does it explicitly state that the code for their methodology is available.
Open Datasets Yes We used images of galaxy of the Galaxy10 dataset [18].
Dataset Splits Yes We generated 2,500 sequences of length τ = 50 with t = 0.05 and separated them into a training, validation, and test sets with 1,000, 500, and 1,000 sequences, respectively.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running its experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers.
Experiment Setup Yes In the experiment with the advection-diffusion dataset reported in Table 1, the selected values of the hyperparameters were α = 0.1, β = 0.01, and γ = 10^6, which were chosen from only eight candidates (see Appendix E for detail).