{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/physnet-a-neural-network-for-predicting","title":"PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments and Partial Charges","arxiv_id":"1902.08408","date":"2019-02-22","proceeding":null,"authors":["Oliver T. Unke","Markus Meuwly"],"abstract":"In recent years, machine learning (ML) methods have become increasingly popular in computational chemistry. After being trained on appropriate ab initio reference data, these methods allow to accurately predict the properties of chemical systems, circumventing the need for explicitly solving the electronic Schr\\\"odinger equation. Because of their computational efficiency and scalability to large datasets, deep neural networks (DNNs) are a particularly promising ML algorithm for chemical applications. This work introduces PhysNet, a DNN architecture designed for predicting energies, forces and dipole moments of chemical systems. PhysNet achieves state-of-the-art performance on the QM9, MD17 and ISO17 benchmarks. Further, two new datasets are generated in order to probe the performance of ML models for describing chemical reactions, long-range interactions, and condensed phase systems. It is shown that explicitly including electrostatics in energy predictions is crucial for a qualitatively correct description of the asymptotic regions of a potential energy surface (PES). PhysNet models trained on a systematically constructed set of small peptide fragments (at most eight heavy atoms) are able to generalize to considerably larger proteins like deca-alanine (Ala$_{10}$): The optimized geometry of helical Ala$_{10}$ predicted by PhysNet is virtually identical to ab initio results (RMSD = 0.21 \\r{A}). By running unbiased molecular dynamics (MD) simulations of Ala$_{10}$ on the PhysNet-PES in gas phase, it is found that instead of a helical structure, Ala$_{10}$ folds into a wreath-shaped configuration, which is more stable than the helical form by 0.46 kcal mol$^{-1}$ according to the reference ab initio calculations.","url_abs":"http://arxiv.org/abs/1902.08408v2","url_pdf":"http://arxiv.org/pdf/1902.08408v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"links_only","authors_date_abstract":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license), from the Kaggle arXiv metadata snapshot of 2026-09-12"},"code_links":[{"paper_slug":"physnet-a-neural-network-for-predicting","repo_url":"https://github.com/MeuwlyGroup/PhysNet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"physnet-a-neural-network-for-predicting","repo_url":"https://github.com/MMunibas/PhysNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"physnet-a-neural-network-for-predicting","repo_url":"https://github.com/MMunibas/fad","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"physnet-a-neural-network-for-predicting","repo_url":"https://github.com/qsb300/PhysNetTF2","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"physnet-a-neural-network-for-predicting","repo_url":"https://github.com/qsb300/PhysNetTF2M","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"physnet-a-neural-network-for-predicting","repo_url":"https://github.com/xiuyu0000/new_papers_codes/tree/main/Cybertron","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/formation-energy-on-qm9","task":"Formation Energy","dataset":"QM9","model":"PhysNet-ens5","rank_in_archive_order":6,"of":18,"metrics":{"MAE":"0.14"},"uses_additional_data":false},{"leaderboard":"/sota/formation-energy-on-qm9","task":"Formation Energy","dataset":"QM9","model":"PhysNet","rank_in_archive_order":10,"of":18,"metrics":{"MAE":"0.19"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1902.08408","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1902.08408"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/qsb300/PhysNetTF2","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/MMunibas/fad","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/MMunibas/PhysNet","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/qsb300/PhysNetTF2M","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/xiuyu0000/new_papers_codes/tree/main/Cybertron","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/MeuwlyGroup/PhysNet","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_honours":1,"unverified":12},"by_repo_kind":{"official":{"samples":7,"ran":1,"repositories":1},"listed":{"samples":6,"ran":0,"repositories":3}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"86d13b9f67274f19","entry":"swish","repo":"MeuwlyGroup/PhysNet","repo_kind":"official","path":"neural_network/activation_fn.py","file_url":"https://github.com/MeuwlyGroup/PhysNet/blob/HEAD/neural_network/activation_fn.py","link_basis":"harvester_set","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"86d13b9f67274f19"}},{"code_sha256_prefix":"09d246ccdbc79091","entry":"semi_orthogonal_glorot_weights","repo":"MeuwlyGroup/PhysNet","repo_kind":"official","path":"neural_network/layers/util.py","file_url":"https://github.com/MeuwlyGroup/PhysNet/blob/HEAD/neural_network/layers/util.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"09d246ccdbc79091"}},{"code_sha256_prefix":"e1a71552a0935d72","entry":"semi_orthogonal_matrix","repo":"MeuwlyGroup/PhysNet","repo_kind":"official","path":"neural_network/layers/util.py","file_url":"https://github.com/MeuwlyGroup/PhysNet/blob/HEAD/neural_network/layers/util.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e1a71552a0935d72"}},{"code_sha256_prefix":"4d8becd4e439e579","entry":"shifted_softplus","repo":"MeuwlyGroup/PhysNet","repo_kind":"official","path":"neural_network/activation_fn.py","file_url":"https://github.com/MeuwlyGroup/PhysNet/blob/HEAD/neural_network/activation_fn.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"4d8becd4e439e579"}},{"code_sha256_prefix":"9660c0db4e564762","entry":"shifted_softplus","repo":"qsb300/PhysNetTF2","repo_kind":"listed","path":"neural_network/activation_fn.py","file_url":"https://github.com/qsb300/PhysNetTF2/blob/HEAD/neural_network/activation_fn.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9660c0db4e564762"}},{"code_sha256_prefix":"1210d8cc78922be5","entry":"shifted_softplus","repo":"qsb300/PhysNetTF2M","repo_kind":"listed","path":"neural_network/activation_fn.py","file_url":"https://github.com/qsb300/PhysNetTF2M/blob/HEAD/neural_network/activation_fn.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"1210d8cc78922be5"}},{"code_sha256_prefix":"c4d6f9c2006dbff2","entry":"softplus","repo":"MeuwlyGroup/PhysNet","repo_kind":"official","path":"neural_network/activation_fn.py","file_url":"https://github.com/MeuwlyGroup/PhysNet/blob/HEAD/neural_network/activation_fn.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c4d6f9c2006dbff2"}},{"code_sha256_prefix":"2afff51165fbb5ea","entry":"softplus","repo":"qsb300/PhysNetTF2","repo_kind":"listed","path":"neural_network/activation_fn.py","file_url":"https://github.com/qsb300/PhysNetTF2/blob/HEAD/neural_network/activation_fn.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"2afff51165fbb5ea"}},{"code_sha256_prefix":"9ac3731380d8f798","entry":"softplus","repo":"qsb300/PhysNetTF2M","repo_kind":"listed","path":"neural_network/activation_fn.py","file_url":"https://github.com/qsb300/PhysNetTF2M/blob/HEAD/neural_network/activation_fn.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9ac3731380d8f798"}},{"code_sha256_prefix":"b898cbbdb9862294","entry":"softplus_inverse","repo":"MeuwlyGroup/PhysNet","repo_kind":"official","path":"neural_network/NeuralNetwork.py","file_url":"https://github.com/MeuwlyGroup/PhysNet/blob/HEAD/neural_network/NeuralNetwork.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b898cbbdb9862294"}},{"code_sha256_prefix":"b924882256391469","entry":"softplus_inverse","repo":"MMunibas/PhysNet","repo_kind":"listed","path":"neural_network/layers/RBFLayer.py","file_url":"https://github.com/MMunibas/PhysNet/blob/HEAD/neural_network/layers/RBFLayer.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b924882256391469"}},{"code_sha256_prefix":"f590f8ed707cb9ab","entry":"square_orthogonal_matrix","repo":"MeuwlyGroup/PhysNet","repo_kind":"official","path":"neural_network/layers/util.py","file_url":"https://github.com/MeuwlyGroup/PhysNet/blob/HEAD/neural_network/layers/util.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f590f8ed707cb9ab"}},{"code_sha256_prefix":"7a3afcf71eb35c83","entry":"swish","repo":"qsb300/PhysNetTF2","repo_kind":"listed","path":"neural_network/activation_fn.py","file_url":"https://github.com/qsb300/PhysNetTF2/blob/HEAD/neural_network/activation_fn.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"7a3afcf71eb35c83"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}