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..
E(n) Equivariant Normalizing Flows
Authors: Victor Garcia Satorras, Emiel Hoogeboom, Fabian Fuchs, Ingmar Posner, Max Welling
NeurIPS 2021 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | 5 Experiments Results: In Table 1 we report the cross-validated Negative Log Likelihood for the test partition. |
| Researcher Affiliation | Academia | Uv A-Bosch Delta Lab, University of Amsterdam1, Department of Engineering Science, University of Oxford2 |
| Pseudocode | No | The paper does not contain any clearly labeled pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide an explicit statement or link regarding the open-sourcing of the code for the methodology described. |
| Open Datasets | Yes | QM9 (Ramakrishnan et al., 2014) is a molecular dataset standarized in machine learning as a chemical property prediction benchmark. |
| Dataset Splits | Yes | For both datasets we use 1,000 validation samples, and 1,000 test samples. |
| Hardware Specification | No | The paper does not specify the exact hardware (e.g., GPU/CPU models, memory) used for running the experiments. |
| Software Dependencies | No | The paper mentions using the 'torchdiffeq package' and 'rdkit toolkit' but does not specify their version numbers or any other software dependencies with versions. |
| Experiment Setup | Yes | Our E-NF method and its non-equivariant variants (GNF, GNF-att, GNF-att-aug) consist of 3 layers each, 32 features per layer, and Si LU activation functions. |