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..
Equivariant Diffusion for Molecule Generation in 3D
Authors: Emiel Hoogeboom, Vı́ctor Garcia Satorras, Clément Vignac, Max Welling
ICML 2022 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimentally, the proposed method significantly outperforms previous 3D molecular generative methods regarding the quality of generated samples and efficiency at training time. |
| Researcher Affiliation | Academia | 1Uv A-Bosch Delta Lab, University of Amsterdam, Netherlands 2EPFL, Lausanne, Switzerland. |
| Pseudocode | Yes | Algorithm 1 Optimizing EDM |
| Open Source Code | No | The paper does not contain an explicit statement or link indicating the release of source code for the described methodology. |
| Open Datasets | Yes | QM9 (Ramakrishnan et al., 2014) is a standard dataset that contains molecular properties and atom coordinates for 130k small molecules with up to 9 heavy atoms (29 atoms including hydrogens). |
| Dataset Splits | Yes | We use the train/val/test partitions introduced in (Anderson et al., 2019), which consists of 100K/18K/13K samples respectively for each partition. |
| Hardware Specification | Yes | Training takes approximately 7 days on a single NVIDIA Ge Force GTX 1080Ti GPU. |
| Software Dependencies | No | The paper mentions software like 'Adam' (optimizer) and 'fully connected neural networks' and general concepts like 'MLPs' and 'EGNN', but does not provide specific version numbers for any software dependencies. |
| Experiment Setup | Yes | All models use 9 layers, 256 features per layer and Si LU activations. They are trained using Adam with batch size 64 and learning rate 10 4. |