{"url":"/sota/unconditional-molecule-generation-on-geom","task":{"name":"Unconditional Molecule Generation","url":"/task/unconditional-molecule-generation","note":null},"dataset":{"name":"GEOM-DRUGS","url":"/dataset/geom-drugs"},"category":"Miscellaneous","categories":["Miscellaneous"],"category_note":null,"description":"This task evaluates the ability of generative models to sample valid and realistic molecular structures.\r\n\r\nThe training dataset can be:\r\n- QM9 (Wu et al., 2018) - consists of 130,000 stable small organic molecules containing up to nine heavy atoms (C, N, O, F) along with hydrogens.\r\n- GEOM-DRUGS (Axelrod and Gómez-Bombarelli, 2022) - consistes of 430,000 large organic molecules of up to 180 atoms. \r\n\r\nFollowing prior work (Hoogeboom et al., 2022), we generally sample 10,000 molecules and compute validity, uniqueness and Posebusters sanity checks (Buttenschoen et al., 2024) for molecules.  Data is generally split following prior work (Hoogeboom et al., 2022, Vignac et al., 2023) to ensure fair comparisons.","description_from":"task","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["PoseBusters Validity","Validity","PoseBusters Atoms Connected"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"PoseBusters Validity":null,"Validity":null,"PoseBusters Atoms Connected":null}},"counts":{"rows":5,"rows_with_code":3,"rows_with_paper_page":5,"rows_dated":5,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"TABASCO","metrics":{"PoseBusters Validity":"92","Validity":"97"},"uses_additional_data":false,"paper_date":"2025-07-01","paper":"/paper/tabasco-a-fast-simplified-model-for-molecular-1","paper_url":"https://arxiv.org/abs/2507.00899v1","paper_title":"TABASCO: A Fast, Simplified Model for Molecular Generation with Improved Physical Quality","code":"https://github.com/carlosinator/tabasco","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"SemlaFlow","metrics":{"PoseBusters Atoms Connected":"92.3","PoseBusters Validity":"87.5","Validity":"93.9"},"uses_additional_data":false,"paper_date":"2024-06-11","paper":"/paper/efficient-3d-molecular-generation-with-flow","paper_url":"https://arxiv.org/abs/2406.07266v3","paper_title":"SemlaFlow -- Efficient 3D Molecular Generation with Latent Attention and Equivariant Flow Matching","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":3,"model":"ADiT","metrics":{"PoseBusters Atoms Connected":"93.0","PoseBusters Validity":"85.3","Validity":"95.3"},"uses_additional_data":false,"paper_date":"2025-03-05","paper":"/paper/all-atom-diffusion-transformers-unified","paper_url":"https://arxiv.org/abs/2503.03965v1","paper_title":"All-atom Diffusion Transformers: Unified generative modelling of molecules and materials","code":"https://github.com/facebookresearch/all-atom-diffusion-transformer","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":1,"n_samples":7,"n_pointer_only_licence":7}},{"rank_in_archive_order":4,"model":"EQGAT-diff","metrics":{"PoseBusters Atoms Connected":"84.4","PoseBusters Validity":"59.7","Validity":"94.6"},"uses_additional_data":false,"paper_date":"2023-09-29","paper":"/paper/navigating-the-design-space-of-equivariant","paper_url":"https://arxiv.org/abs/2309.17296v2","paper_title":"Navigating the Design Space of Equivariant Diffusion-Based Generative Models for De Novo 3D Molecule Generation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":5,"model":"MiDi","metrics":{"Validity":"77.8"},"uses_additional_data":false,"paper_date":"2023-02-17","paper":"/paper/midi-mixed-graph-and-3d-denoising-diffusion","paper_url":"https://arxiv.org/abs/2302.09048v2","paper_title":"MiDi: Mixed Graph and 3D Denoising Diffusion for Molecule Generation","code":"https://github.com/cvignac/midi","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":11,"n_samples":13,"n_pointer_only_licence":0}}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 6,264 of the 9,581 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9581,"papers_checked":6264,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":3316},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-24T18:15:14+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":2,"rows_with_any_sample_ran":2,"distinct_papers_with_graph_line":2,"distinct_papers_with_any_sample_ran":2,"samples_over_distinct_papers":{"n_ran":8,"n_unverified":12,"n_samples":20,"n_pointer_only_licence":7,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":8,"n_unverified":12,"n_samples":20,"n_pointer_only_licence":7,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}