Metraipolis › archive › dc_e8e372cbf265
Geometric Relevance Is Not Sufficient: Evaluating Focused Post-Training Supervision in Diffusion Docking — arXiv:2611.08249
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.08249
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-06T14:50:55.621Z
- SHA-256 of the archived text
40bf420033acaf1fb4fe6969d964e7e24fccf8b195df3dedc29560a8e4ae6011- Stable URL of this page
https://metraipolis.com/documento/dc_e8e372cbf265
How to cite
Geometric Relevance Is Not Sufficient: Evaluating Focused Post-Training Supervision in Diffusion Docking — arXiv:2611.08249 — arXiv (en), https://arxiv.org/abs/2611.08249 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e8e372cbf265 (Metraipolis archive dc_e8e372cbf265, sha256 40bf420033acaf1fb4fe6969d964e7e24fccf8b195df3dedc29560a8e4ae6011)
BibTeX
@misc{dce8e372cbf265,
title = {Geometric Relevance Is Not Sufficient: Evaluating Focused Post-Training Supervision in Diffusion Docking — arXiv:2611.08249},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.08249},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e8e372cbf265; sha256 40bf420033acaf1fb4fe6969d964e7e24fccf8b195df3dedc29560a8e4ae6011},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.08249. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Diffusion-based molecular docking requires learning a corrective score field across a broad corruption distribution, yet individual pose errors can differ substantially in their structural consequences. This creates a post-training conflict between preserving broad score-field coverage and concentrating supervision on corrections that appear more relevant to docking geometry. Using DiffDock, a fixed-budget study compares broad continuation with torsional supervision focused on Cartesian consequence or conditional target consistency, and with a broad/focused mixture. Evaluation separates fixed score fidelity, generated-pose quality, best-of-10 sampling capacity, and confidence-selected Top-1 docking performance. At <5 Angstrom, broad continuation had the highest mean Top-1 success among trained policies, and focused or mixed supervision showed no consistent advantage under the tested budget. The starting model had better fixed score fidelity, whereas the evaluated broad-continuation model had higher mean Top-1 success with the same aggregate best-of-10 recovery. Focusing supervision on structurally consequential or more consistently specified corrections did not consistently improve returned-pose success over broad continuation. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Biomolecules (q-bio.BM). Pagina del lavoro: https://arxiv.org/abs/2611.08249 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-11-06T14:21:42.645Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).