{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:64JDHNSZB4URZVOS3POXNPJ7MQ","short_pith_number":"pith:64JDHNSZ","schema_version":"1.0","canonical_sha256":"f71233b6590f291cd5d2dbdd76bd3f6407c591b0caaf861208f2e31222eb1e00","source":{"kind":"arxiv","id":"2606.12651","version":1},"attestation_state":"computed","paper":{"title":"Physics-Aware Auxiliary Losses Improve Out-of-Distribution Generalization of a GNN Synthesizability Filter","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["q-bio.QM"],"primary_cat":"cs.LG","authors_text":"Dhruv Agarwal, Riya Bisht","submitted_at":"2026-06-10T20:21:56Z","abstract_excerpt":"Machine-learning drug-discovery pipelines increasingly rely on generative models that propose molecules far from the data used to train downstream synthesizability filters. Existing\n  filters (SAScore, SCScore, RAscore, DeepSA) are purely statistical and degrade in exactly this out-of-distribution (OOD) regime. We ask whether cheap, closed-form physical priors, used\n  as auxiliary supervision on a graph neural network (GNN), improve OOD generalization. We add two auxiliary losses to a GINE backbone: a topological complexity regression supervised by\n  the Bertz index, and a strain-energy soft p"},"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":"2606.12651","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-10T20:21:56Z","cross_cats_sorted":["q-bio.QM"],"title_canon_sha256":"c7b0cd08091d9b9c788c90ca90719e79d174a3812e3608f3b3becff9b090e5fd","abstract_canon_sha256":"13873bdfb01cdfb3e543c7ca9bc3ac3c36ae708e25511fff28ec866103f4c299"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-12T01:08:42.735405Z","signature_b64":"6hRf8Uom7oJuOGWlmjfTgCKNo09Qv6COpxaIqq3Rzijg1Ns5pyTy7xpEGOyxYD7Z6v2xa3QI6btKoVPZIACMBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f71233b6590f291cd5d2dbdd76bd3f6407c591b0caaf861208f2e31222eb1e00","last_reissued_at":"2026-06-12T01:08:42.734405Z","signature_status":"signed_v1","first_computed_at":"2026-06-12T01:08:42.734405Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Physics-Aware Auxiliary Losses Improve Out-of-Distribution Generalization of a GNN Synthesizability Filter","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["q-bio.QM"],"primary_cat":"cs.LG","authors_text":"Dhruv Agarwal, Riya Bisht","submitted_at":"2026-06-10T20:21:56Z","abstract_excerpt":"Machine-learning drug-discovery pipelines increasingly rely on generative models that propose molecules far from the data used to train downstream synthesizability filters. Existing\n  filters (SAScore, SCScore, RAscore, DeepSA) are purely statistical and degrade in exactly this out-of-distribution (OOD) regime. We ask whether cheap, closed-form physical priors, used\n  as auxiliary supervision on a graph neural network (GNN), improve OOD generalization. We add two auxiliary losses to a GINE backbone: a topological complexity regression supervised by\n  the Bertz index, and a strain-energy soft p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.12651","kind":"arxiv","version":1},"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/2606.12651/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":"2606.12651","created_at":"2026-06-12T01:08:42.734560+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.12651v1","created_at":"2026-06-12T01:08:42.734560+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.12651","created_at":"2026-06-12T01:08:42.734560+00:00"},{"alias_kind":"pith_short_12","alias_value":"64JDHNSZB4UR","created_at":"2026-06-12T01:08:42.734560+00:00"},{"alias_kind":"pith_short_16","alias_value":"64JDHNSZB4URZVOS","created_at":"2026-06-12T01:08:42.734560+00:00"},{"alias_kind":"pith_short_8","alias_value":"64JDHNSZ","created_at":"2026-06-12T01:08:42.734560+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/64JDHNSZB4URZVOS3POXNPJ7MQ","json":"https://pith.science/pith/64JDHNSZB4URZVOS3POXNPJ7MQ.json","graph_json":"https://pith.science/api/pith-number/64JDHNSZB4URZVOS3POXNPJ7MQ/graph.json","events_json":"https://pith.science/api/pith-number/64JDHNSZB4URZVOS3POXNPJ7MQ/events.json","paper":"https://pith.science/paper/64JDHNSZ"},"agent_actions":{"view_html":"https://pith.science/pith/64JDHNSZB4URZVOS3POXNPJ7MQ","download_json":"https://pith.science/pith/64JDHNSZB4URZVOS3POXNPJ7MQ.json","view_paper":"https://pith.science/paper/64JDHNSZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.12651&json=true","fetch_graph":"https://pith.science/api/pith-number/64JDHNSZB4URZVOS3POXNPJ7MQ/graph.json","fetch_events":"https://pith.science/api/pith-number/64JDHNSZB4URZVOS3POXNPJ7MQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/64JDHNSZB4URZVOS3POXNPJ7MQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/64JDHNSZB4URZVOS3POXNPJ7MQ/action/storage_attestation","attest_author":"https://pith.science/pith/64JDHNSZB4URZVOS3POXNPJ7MQ/action/author_attestation","sign_citation":"https://pith.science/pith/64JDHNSZB4URZVOS3POXNPJ7MQ/action/citation_signature","submit_replication":"https://pith.science/pith/64JDHNSZB4URZVOS3POXNPJ7MQ/action/replication_record"}},"created_at":"2026-06-12T01:08:42.734560+00:00","updated_at":"2026-06-12T01:08:42.734560+00:00"}