{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:GGULXS7OXWP4CCSOAS4RSEHMO7","short_pith_number":"pith:GGULXS7O","schema_version":"1.0","canonical_sha256":"31a8bbcbeebd9fc10a4e04b91910ec77cd67ec0f71414fb969e6e1fbb2f76b1e","source":{"kind":"arxiv","id":"2508.13408","version":2},"attestation_state":"computed","paper":{"title":"NovoMolGen: Rethinking Molecular Language Model Pretraining","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Kamran Chitsaz, Nirav Pravinbhai Bhatt, Quentin Fournier, Roshan Balaji, Sarath Chandar","submitted_at":"2025-08-19T00:04:48Z","abstract_excerpt":"Designing de-novo molecules with desired property profiles requires efficient exploration of the vast chemical space ranging from $10^{23}$ to $10^{60}$ possible synthesizable candidates. While various deep generative models have been developed to design small molecules using diverse input representations, Molecular Large Language Models (Mol-LLMs) based on string representations have emerged as a scalable approach capable of exploring billions of molecules. However, there remains limited understanding regarding how standard language modeling practices such as textual representations, tokeniza"},"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":"2508.13408","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-08-19T00:04:48Z","cross_cats_sorted":[],"title_canon_sha256":"e5c7ae83651b65ae3bbe218bae614d0e839bec2e06e81942e9e41fca6299c4fa","abstract_canon_sha256":"c307a54b842f31ce0074bc287f7ec61c7174acba1c2bf0be197dfb2912a67b83"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:57:39.014299Z","signature_b64":"oYdOlhvd7tZ8iw0NJ1Din1LY1tHyjXdvaYjURFJ25iIdXB4zuMQTaGPohxZ4RB1cjOnV6EFph5HCoqrmF+XzDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"31a8bbcbeebd9fc10a4e04b91910ec77cd67ec0f71414fb969e6e1fbb2f76b1e","last_reissued_at":"2026-07-05T11:57:39.013847Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:57:39.013847Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"NovoMolGen: Rethinking Molecular Language Model Pretraining","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Kamran Chitsaz, Nirav Pravinbhai Bhatt, Quentin Fournier, Roshan Balaji, Sarath Chandar","submitted_at":"2025-08-19T00:04:48Z","abstract_excerpt":"Designing de-novo molecules with desired property profiles requires efficient exploration of the vast chemical space ranging from $10^{23}$ to $10^{60}$ possible synthesizable candidates. While various deep generative models have been developed to design small molecules using diverse input representations, Molecular Large Language Models (Mol-LLMs) based on string representations have emerged as a scalable approach capable of exploring billions of molecules. However, there remains limited understanding regarding how standard language modeling practices such as textual representations, tokeniza"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.13408","kind":"arxiv","version":2},"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/2508.13408/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":"2508.13408","created_at":"2026-07-05T11:57:39.013895+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.13408v2","created_at":"2026-07-05T11:57:39.013895+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.13408","created_at":"2026-07-05T11:57:39.013895+00:00"},{"alias_kind":"pith_short_12","alias_value":"GGULXS7OXWP4","created_at":"2026-07-05T11:57:39.013895+00:00"},{"alias_kind":"pith_short_16","alias_value":"GGULXS7OXWP4CCSO","created_at":"2026-07-05T11:57:39.013895+00:00"},{"alias_kind":"pith_short_8","alias_value":"GGULXS7O","created_at":"2026-07-05T11:57:39.013895+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.12113","citing_title":"Augmenting Molecular Language Models with Local $n$-gram Memory","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10230","citing_title":"FORGE: Fragment-Oriented Ranking and Generation for Context-Aware Molecular Optimization","ref_index":37,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GGULXS7OXWP4CCSOAS4RSEHMO7","json":"https://pith.science/pith/GGULXS7OXWP4CCSOAS4RSEHMO7.json","graph_json":"https://pith.science/api/pith-number/GGULXS7OXWP4CCSOAS4RSEHMO7/graph.json","events_json":"https://pith.science/api/pith-number/GGULXS7OXWP4CCSOAS4RSEHMO7/events.json","paper":"https://pith.science/paper/GGULXS7O"},"agent_actions":{"view_html":"https://pith.science/pith/GGULXS7OXWP4CCSOAS4RSEHMO7","download_json":"https://pith.science/pith/GGULXS7OXWP4CCSOAS4RSEHMO7.json","view_paper":"https://pith.science/paper/GGULXS7O","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.13408&json=true","fetch_graph":"https://pith.science/api/pith-number/GGULXS7OXWP4CCSOAS4RSEHMO7/graph.json","fetch_events":"https://pith.science/api/pith-number/GGULXS7OXWP4CCSOAS4RSEHMO7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GGULXS7OXWP4CCSOAS4RSEHMO7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GGULXS7OXWP4CCSOAS4RSEHMO7/action/storage_attestation","attest_author":"https://pith.science/pith/GGULXS7OXWP4CCSOAS4RSEHMO7/action/author_attestation","sign_citation":"https://pith.science/pith/GGULXS7OXWP4CCSOAS4RSEHMO7/action/citation_signature","submit_replication":"https://pith.science/pith/GGULXS7OXWP4CCSOAS4RSEHMO7/action/replication_record"}},"created_at":"2026-07-05T11:57:39.013895+00:00","updated_at":"2026-07-05T11:57:39.013895+00:00"}