{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:VGBZDYT55N77GFRVQUV43SFD4D","short_pith_number":"pith:VGBZDYT5","schema_version":"1.0","canonical_sha256":"a98391e27deb7ff31635852bcdc8a3e0c67d42e180892d3bd3143c030a745995","source":{"kind":"arxiv","id":"2501.05550","version":2},"attestation_state":"computed","paper":{"title":"Emergent weight morphologies in deep neural networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cond-mat.dis-nn"],"primary_cat":"cs.LG","authors_text":"Felix Meigel, Pascal de Jong, Steffen Rulands","submitted_at":"2025-01-09T19:48:51Z","abstract_excerpt":"Whether deep neural networks can exhibit emergent behaviour is not only relevant for understanding how deep learning works, it is also pivotal for estimating potential security risks of increasingly capable artificial intelligence systems. Here, we show that training deep neural networks gives rise to emergent weight morphologies independent of the training data. Specifically, in analogy to condensed matter physics, we derive a theory that predict that the homogeneous state of deep neural networks is unstable in a way that leads to the emergence of periodic channel structures. We verified thes"},"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":"2501.05550","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-01-09T19:48:51Z","cross_cats_sorted":["cond-mat.dis-nn"],"title_canon_sha256":"d8684bcb8f3f4b905d579e90899af724355d59f3f2ed54d52306a253e130a565","abstract_canon_sha256":"2f3a72c76d6c83dbbe753bc387f039009f2c8ca1994720842c748d4374175bb6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:46:59.787214Z","signature_b64":"ohS2HDsv4wn0j2gxSRJfyZhMJGoDF8QVoBWrxrkB8aJqaB1JfaL4nNzQQ2utt1cDF8pcZ/MGJcmJqD+7e4ipBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a98391e27deb7ff31635852bcdc8a3e0c67d42e180892d3bd3143c030a745995","last_reissued_at":"2026-07-05T10:46:59.786660Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:46:59.786660Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Emergent weight morphologies in deep neural networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cond-mat.dis-nn"],"primary_cat":"cs.LG","authors_text":"Felix Meigel, Pascal de Jong, Steffen Rulands","submitted_at":"2025-01-09T19:48:51Z","abstract_excerpt":"Whether deep neural networks can exhibit emergent behaviour is not only relevant for understanding how deep learning works, it is also pivotal for estimating potential security risks of increasingly capable artificial intelligence systems. Here, we show that training deep neural networks gives rise to emergent weight morphologies independent of the training data. Specifically, in analogy to condensed matter physics, we derive a theory that predict that the homogeneous state of deep neural networks is unstable in a way that leads to the emergence of periodic channel structures. We verified thes"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.05550","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/2501.05550/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":"2501.05550","created_at":"2026-07-05T10:46:59.786718+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.05550v2","created_at":"2026-07-05T10:46:59.786718+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.05550","created_at":"2026-07-05T10:46:59.786718+00:00"},{"alias_kind":"pith_short_12","alias_value":"VGBZDYT55N77","created_at":"2026-07-05T10:46:59.786718+00:00"},{"alias_kind":"pith_short_16","alias_value":"VGBZDYT55N77GFRV","created_at":"2026-07-05T10:46:59.786718+00:00"},{"alias_kind":"pith_short_8","alias_value":"VGBZDYT5","created_at":"2026-07-05T10:46:59.786718+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/VGBZDYT55N77GFRVQUV43SFD4D","json":"https://pith.science/pith/VGBZDYT55N77GFRVQUV43SFD4D.json","graph_json":"https://pith.science/api/pith-number/VGBZDYT55N77GFRVQUV43SFD4D/graph.json","events_json":"https://pith.science/api/pith-number/VGBZDYT55N77GFRVQUV43SFD4D/events.json","paper":"https://pith.science/paper/VGBZDYT5"},"agent_actions":{"view_html":"https://pith.science/pith/VGBZDYT55N77GFRVQUV43SFD4D","download_json":"https://pith.science/pith/VGBZDYT55N77GFRVQUV43SFD4D.json","view_paper":"https://pith.science/paper/VGBZDYT5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.05550&json=true","fetch_graph":"https://pith.science/api/pith-number/VGBZDYT55N77GFRVQUV43SFD4D/graph.json","fetch_events":"https://pith.science/api/pith-number/VGBZDYT55N77GFRVQUV43SFD4D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VGBZDYT55N77GFRVQUV43SFD4D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VGBZDYT55N77GFRVQUV43SFD4D/action/storage_attestation","attest_author":"https://pith.science/pith/VGBZDYT55N77GFRVQUV43SFD4D/action/author_attestation","sign_citation":"https://pith.science/pith/VGBZDYT55N77GFRVQUV43SFD4D/action/citation_signature","submit_replication":"https://pith.science/pith/VGBZDYT55N77GFRVQUV43SFD4D/action/replication_record"}},"created_at":"2026-07-05T10:46:59.786718+00:00","updated_at":"2026-07-05T10:46:59.786718+00:00"}