{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:YSEIRDAUCHMJSJYKKEGHWNECTL","short_pith_number":"pith:YSEIRDAU","schema_version":"1.0","canonical_sha256":"c488888c1411d899270a510c7b34829af626d0b637183ceff433497e95239198","source":{"kind":"arxiv","id":"2306.07104","version":1},"attestation_state":"computed","paper":{"title":"Unveiling the Hessian's Connection to the Decision Boundary","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cond-mat.dis-nn","stat.ML"],"primary_cat":"cs.LG","authors_text":"Anna Dawid, Freya Behrens, Mahalakshmi Sabanayagam, Urte Adomaityte","submitted_at":"2023-06-12T13:27:55Z","abstract_excerpt":"Understanding the properties of well-generalizing minima is at the heart of deep learning research. On the one hand, the generalization of neural networks has been connected to the decision boundary complexity, which is hard to study in the high-dimensional input space. Conversely, the flatness of a minimum has become a controversial proxy for generalization. In this work, we provide the missing link between the two approaches and show that the Hessian top eigenvectors characterize the decision boundary learned by the neural network. Notably, the number of outliers in the Hessian spectrum is 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":"2306.07104","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-06-12T13:27:55Z","cross_cats_sorted":["cond-mat.dis-nn","stat.ML"],"title_canon_sha256":"d548a103dcf621f89239ab2ca1215cdcb510beeae15450c674508956b82bada1","abstract_canon_sha256":"1d8f9081118e8149dcb9900b88559ad0dc3ca94c9e74c99aa9f00fbb6074d679"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:19:47.725586Z","signature_b64":"Rmzq27aMfJ9iop8t/H+dkaj6uWrMGmhB7CVx9oEM8UQZrwBxubotBEhTtqr6MO7hW7hdCdyaOzHuLISRMDZSCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c488888c1411d899270a510c7b34829af626d0b637183ceff433497e95239198","last_reissued_at":"2026-07-05T06:19:47.725177Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:19:47.725177Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Unveiling the Hessian's Connection to the Decision Boundary","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cond-mat.dis-nn","stat.ML"],"primary_cat":"cs.LG","authors_text":"Anna Dawid, Freya Behrens, Mahalakshmi Sabanayagam, Urte Adomaityte","submitted_at":"2023-06-12T13:27:55Z","abstract_excerpt":"Understanding the properties of well-generalizing minima is at the heart of deep learning research. On the one hand, the generalization of neural networks has been connected to the decision boundary complexity, which is hard to study in the high-dimensional input space. Conversely, the flatness of a minimum has become a controversial proxy for generalization. In this work, we provide the missing link between the two approaches and show that the Hessian top eigenvectors characterize the decision boundary learned by the neural network. Notably, the number of outliers in the Hessian spectrum is p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.07104","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/2306.07104/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":"2306.07104","created_at":"2026-07-05T06:19:47.725236+00:00"},{"alias_kind":"arxiv_version","alias_value":"2306.07104v1","created_at":"2026-07-05T06:19:47.725236+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.07104","created_at":"2026-07-05T06:19:47.725236+00:00"},{"alias_kind":"pith_short_12","alias_value":"YSEIRDAUCHMJ","created_at":"2026-07-05T06:19:47.725236+00:00"},{"alias_kind":"pith_short_16","alias_value":"YSEIRDAUCHMJSJYK","created_at":"2026-07-05T06:19:47.725236+00:00"},{"alias_kind":"pith_short_8","alias_value":"YSEIRDAU","created_at":"2026-07-05T06:19:47.725236+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.19356","citing_title":"Microscopic imprints of learned solutions in adaptive resistor networks","ref_index":5,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YSEIRDAUCHMJSJYKKEGHWNECTL","json":"https://pith.science/pith/YSEIRDAUCHMJSJYKKEGHWNECTL.json","graph_json":"https://pith.science/api/pith-number/YSEIRDAUCHMJSJYKKEGHWNECTL/graph.json","events_json":"https://pith.science/api/pith-number/YSEIRDAUCHMJSJYKKEGHWNECTL/events.json","paper":"https://pith.science/paper/YSEIRDAU"},"agent_actions":{"view_html":"https://pith.science/pith/YSEIRDAUCHMJSJYKKEGHWNECTL","download_json":"https://pith.science/pith/YSEIRDAUCHMJSJYKKEGHWNECTL.json","view_paper":"https://pith.science/paper/YSEIRDAU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2306.07104&json=true","fetch_graph":"https://pith.science/api/pith-number/YSEIRDAUCHMJSJYKKEGHWNECTL/graph.json","fetch_events":"https://pith.science/api/pith-number/YSEIRDAUCHMJSJYKKEGHWNECTL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YSEIRDAUCHMJSJYKKEGHWNECTL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YSEIRDAUCHMJSJYKKEGHWNECTL/action/storage_attestation","attest_author":"https://pith.science/pith/YSEIRDAUCHMJSJYKKEGHWNECTL/action/author_attestation","sign_citation":"https://pith.science/pith/YSEIRDAUCHMJSJYKKEGHWNECTL/action/citation_signature","submit_replication":"https://pith.science/pith/YSEIRDAUCHMJSJYKKEGHWNECTL/action/replication_record"}},"created_at":"2026-07-05T06:19:47.725236+00:00","updated_at":"2026-07-05T06:19:47.725236+00:00"}