{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:GHCHMELNAYOFALIGFQ4A7EXIRO","short_pith_number":"pith:GHCHMELN","schema_version":"1.0","canonical_sha256":"31c476116d061c502d062c380f92e88b8780a44d8ce99802a01063cac00fb2ec","source":{"kind":"arxiv","id":"2506.07579","version":1},"attestation_state":"computed","paper":{"title":"Beyond Scaling: Chemical Intuition as Emergent Ability of Universal Machine Learning Interatomic Potentials","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cond-mat.mtrl-sci","authors_text":"Aiichiro Nakano, Ken-ichi Nomura, Kohei Shimamura, Priya Vashishta, Rajiv K. Kalia, Shinnosuke Hattori","submitted_at":"2025-06-09T09:23:15Z","abstract_excerpt":"Machine Learning Interatomic Potentials (MLIPs) have successfully demonstrated scaling behavior, i.e. the power-law improvement in training performance, however the emergence of novel capabilities at scale remains unexplored. We have developed Edge-wise Emergent Decomposition (E3D) framework to investigate how an MLIP develops the ability to derive physically meaningful local representations of chemical bonds without explicit supervision. Employing an E(3)-equivariant network (Allegro) trained on molecular data (SPICE~2), we found that the trained MLIP spontaneously learned representations of "},"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":"2506.07579","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cond-mat.mtrl-sci","submitted_at":"2025-06-09T09:23:15Z","cross_cats_sorted":[],"title_canon_sha256":"eda58dbabf3303ed8c5d5b56606114d04dbb6164de403fe937fb344d4ce9d7a3","abstract_canon_sha256":"66590a893d6deebf27d0c721f8b1d5dda6ba002fb662deb80109f53bbfaed5e8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:18:29.748031Z","signature_b64":"GK2TLBuneuo9UD3kY57WXWy0miRL2TgnA/uu9C0ZCMO6JOWLLA+7X2hTCix+zHSXwZrRFpgwAqkqwK8BbmzMBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"31c476116d061c502d062c380f92e88b8780a44d8ce99802a01063cac00fb2ec","last_reissued_at":"2026-07-05T11:18:29.747563Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:18:29.747563Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Scaling: Chemical Intuition as Emergent Ability of Universal Machine Learning Interatomic Potentials","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cond-mat.mtrl-sci","authors_text":"Aiichiro Nakano, Ken-ichi Nomura, Kohei Shimamura, Priya Vashishta, Rajiv K. Kalia, Shinnosuke Hattori","submitted_at":"2025-06-09T09:23:15Z","abstract_excerpt":"Machine Learning Interatomic Potentials (MLIPs) have successfully demonstrated scaling behavior, i.e. the power-law improvement in training performance, however the emergence of novel capabilities at scale remains unexplored. We have developed Edge-wise Emergent Decomposition (E3D) framework to investigate how an MLIP develops the ability to derive physically meaningful local representations of chemical bonds without explicit supervision. Employing an E(3)-equivariant network (Allegro) trained on molecular data (SPICE~2), we found that the trained MLIP spontaneously learned representations of "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.07579","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/2506.07579/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":"2506.07579","created_at":"2026-07-05T11:18:29.747614+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.07579v1","created_at":"2026-07-05T11:18:29.747614+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.07579","created_at":"2026-07-05T11:18:29.747614+00:00"},{"alias_kind":"pith_short_12","alias_value":"GHCHMELNAYOF","created_at":"2026-07-05T11:18:29.747614+00:00"},{"alias_kind":"pith_short_16","alias_value":"GHCHMELNAYOFALIG","created_at":"2026-07-05T11:18:29.747614+00:00"},{"alias_kind":"pith_short_8","alias_value":"GHCHMELN","created_at":"2026-07-05T11:18:29.747614+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.29100","citing_title":"Toward Exascale AI for Science: A Scalable AI Skill for Autonomous Microkinetics Discovery","ref_index":18,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GHCHMELNAYOFALIGFQ4A7EXIRO","json":"https://pith.science/pith/GHCHMELNAYOFALIGFQ4A7EXIRO.json","graph_json":"https://pith.science/api/pith-number/GHCHMELNAYOFALIGFQ4A7EXIRO/graph.json","events_json":"https://pith.science/api/pith-number/GHCHMELNAYOFALIGFQ4A7EXIRO/events.json","paper":"https://pith.science/paper/GHCHMELN"},"agent_actions":{"view_html":"https://pith.science/pith/GHCHMELNAYOFALIGFQ4A7EXIRO","download_json":"https://pith.science/pith/GHCHMELNAYOFALIGFQ4A7EXIRO.json","view_paper":"https://pith.science/paper/GHCHMELN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.07579&json=true","fetch_graph":"https://pith.science/api/pith-number/GHCHMELNAYOFALIGFQ4A7EXIRO/graph.json","fetch_events":"https://pith.science/api/pith-number/GHCHMELNAYOFALIGFQ4A7EXIRO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GHCHMELNAYOFALIGFQ4A7EXIRO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GHCHMELNAYOFALIGFQ4A7EXIRO/action/storage_attestation","attest_author":"https://pith.science/pith/GHCHMELNAYOFALIGFQ4A7EXIRO/action/author_attestation","sign_citation":"https://pith.science/pith/GHCHMELNAYOFALIGFQ4A7EXIRO/action/citation_signature","submit_replication":"https://pith.science/pith/GHCHMELNAYOFALIGFQ4A7EXIRO/action/replication_record"}},"created_at":"2026-07-05T11:18:29.747614+00:00","updated_at":"2026-07-05T11:18:29.747614+00:00"}