{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:KDF6MNBYJXAK5UQZBWCWWDFB4B","short_pith_number":"pith:KDF6MNBY","schema_version":"1.0","canonical_sha256":"50cbe634384dc0aed2190d856b0ca1e07f3c9c600a3b2ab8c63b4c5d995da475","source":{"kind":"arxiv","id":"2110.04126","version":4},"attestation_state":"computed","paper":{"title":"3D Infomax improves GNNs for Molecular Property Prediction","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","q-bio.BM"],"primary_cat":"cs.LG","authors_text":"Christian Dallago, Dominique Beaini, Gabriele Corso, Hannes St\\\"ark, Pietro Li\\`o, Prudencio Tossou, Stephan G\\\"unnemann","submitted_at":"2021-10-08T13:30:49Z","abstract_excerpt":"Molecular property prediction is one of the fastest-growing applications of deep learning with critical real-world impacts. Including 3D molecular structure as input to learned models improves their performance for many molecular tasks. However, this information is infeasible to compute at the scale required by several real-world applications. We propose pre-training a model to reason about the geometry of molecules given only their 2D molecular graphs. Using methods from self-supervised learning, we maximize the mutual information between 3D summary vectors and the representations of a Graph "},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2110.04126","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-10-08T13:30:49Z","cross_cats_sorted":["cs.AI","q-bio.BM"],"title_canon_sha256":"c3e0fa3ecd4f32b4277e334d829cdb160a92c140f6cb26664d51d074d8e4c35f","abstract_canon_sha256":"078c87ac5f9085f65e9cdf4bfbe3842b9aa9eec06f7b007ba1a469a74421c1ae"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:28:52.868186Z","signature_b64":"uLTN1Zd76xWXBqVh4UgjDQe0fnqVfx5BTSPkafKFYTAVFjqCIOxaSNUybwCNsBZIGlPp6Nue++nc/xicDy7BAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"50cbe634384dc0aed2190d856b0ca1e07f3c9c600a3b2ab8c63b4c5d995da475","last_reissued_at":"2026-07-05T04:28:52.867666Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:28:52.867666Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"3D Infomax improves GNNs for Molecular Property Prediction","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","q-bio.BM"],"primary_cat":"cs.LG","authors_text":"Christian Dallago, Dominique Beaini, Gabriele Corso, Hannes St\\\"ark, Pietro Li\\`o, Prudencio Tossou, Stephan G\\\"unnemann","submitted_at":"2021-10-08T13:30:49Z","abstract_excerpt":"Molecular property prediction is one of the fastest-growing applications of deep learning with critical real-world impacts. Including 3D molecular structure as input to learned models improves their performance for many molecular tasks. However, this information is infeasible to compute at the scale required by several real-world applications. We propose pre-training a model to reason about the geometry of molecules given only their 2D molecular graphs. Using methods from self-supervised learning, we maximize the mutual information between 3D summary vectors and the representations of a Graph "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2110.04126","kind":"arxiv","version":4},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2110.04126/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2110.04126","created_at":"2026-07-05T04:28:52.867727+00:00"},{"alias_kind":"arxiv_version","alias_value":"2110.04126v4","created_at":"2026-07-05T04:28:52.867727+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2110.04126","created_at":"2026-07-05T04:28:52.867727+00:00"},{"alias_kind":"pith_short_12","alias_value":"KDF6MNBYJXAK","created_at":"2026-07-05T04:28:52.867727+00:00"},{"alias_kind":"pith_short_16","alias_value":"KDF6MNBYJXAK5UQZ","created_at":"2026-07-05T04:28:52.867727+00:00"},{"alias_kind":"pith_short_8","alias_value":"KDF6MNBY","created_at":"2026-07-05T04:28:52.867727+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.02419","citing_title":"DPA4: Pushing the Accuracy-Cost Frontier of Interatomic Potentials with EMFA SO(2) Convolution","ref_index":174,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08679","citing_title":"Attention-based graph neural networks: a survey","ref_index":170,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KDF6MNBYJXAK5UQZBWCWWDFB4B","json":"https://pith.science/pith/KDF6MNBYJXAK5UQZBWCWWDFB4B.json","graph_json":"https://pith.science/api/pith-number/KDF6MNBYJXAK5UQZBWCWWDFB4B/graph.json","events_json":"https://pith.science/api/pith-number/KDF6MNBYJXAK5UQZBWCWWDFB4B/events.json","paper":"https://pith.science/paper/KDF6MNBY"},"agent_actions":{"view_html":"https://pith.science/pith/KDF6MNBYJXAK5UQZBWCWWDFB4B","download_json":"https://pith.science/pith/KDF6MNBYJXAK5UQZBWCWWDFB4B.json","view_paper":"https://pith.science/paper/KDF6MNBY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2110.04126&json=true","fetch_graph":"https://pith.science/api/pith-number/KDF6MNBYJXAK5UQZBWCWWDFB4B/graph.json","fetch_events":"https://pith.science/api/pith-number/KDF6MNBYJXAK5UQZBWCWWDFB4B/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KDF6MNBYJXAK5UQZBWCWWDFB4B/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KDF6MNBYJXAK5UQZBWCWWDFB4B/action/storage_attestation","attest_author":"https://pith.science/pith/KDF6MNBYJXAK5UQZBWCWWDFB4B/action/author_attestation","sign_citation":"https://pith.science/pith/KDF6MNBYJXAK5UQZBWCWWDFB4B/action/citation_signature","submit_replication":"https://pith.science/pith/KDF6MNBYJXAK5UQZBWCWWDFB4B/action/replication_record"}},"created_at":"2026-07-05T04:28:52.867727+00:00","updated_at":"2026-07-05T04:28:52.867727+00:00"}