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].
Controlled Generation with Equivariant Variational Flow Matching
Authors: Floor Eijkelboom, Heiko Zimmermann, Sharvaree Vadgama, Erik J Bekkers, Max Welling, Christian A. Naesseth, Jan-Willem Van De Meent
ICML 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We evaluate our approach on both uncontrolled and controlled molecular generation, achieving state-of-the-art performance on uncontrolled generation and outperforming stateof-the-art models in controlled generation, both with end-to-end training and in the Bayesian inference setting. This work strengthens the connection between flow-based generative modeling and Bayesian inference, offering a scalable and principled framework for constraint-driven and symmetry-aware generation. |
| Researcher Affiliation | Collaboration | Floor Eijkelboom 1 Heiko Zimmermann 2 Sharvaree Vadgama 2 Erik J Bekkers 2 Max Welling 1 Christian A. Naesseth 1 Jan-Willem van de Meent 1 ... 1Bosch-Delta Lab 2AMLab. Correspondence to: Floor Eijkelboom <EMAIL>. |
| Pseudocode | Yes | Algorithm 1 Sampling Controlled VFM |
| Open Source Code | No | The paper does not contain any explicit statements about releasing source code for the described methodology, nor does it provide a link to a code repository. |
| Open Datasets | Yes | We evaluate GVFM on the QM9 dataset using established metrics... Our results in Table 1 confirm strong performance across both QM9 and ZINC250k datasets... We evaluate G-VFM s capabilities for joint molecular generation on the QM9 and GEOM-Drugs datasets |
| Dataset Splits | No | The paper mentions using well-known datasets like QM9, ZINC250k, and GEOM-Drugs, but it does not explicitly state the training, validation, or test split percentages or sample counts for any of these datasets in the main text. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU/CPU models, memory specifications) used for running the experiments. |
| Software Dependencies | No | The paper mentions RDKit (Landrum, 2016) for parsing validity but does not specify a version number for RDKit or any other software dependencies like programming languages or libraries. |
| Experiment Setup | No | The paper describes the methodology and evaluation, but does not provide specific experimental setup details such as hyperparameter values (e.g., learning rate, batch size, optimizer settings) in the main text. It refers to 'Property classifiers are trained following (Hoogeboom et al., 2022a).' |