{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:2PKVW5PS5YICT5MUPVYETVZJI3","short_pith_number":"pith:2PKVW5PS","schema_version":"1.0","canonical_sha256":"d3d55b75f2ee1029f5947d7049d72946ca60e9c6ae720ef76fc9b1db7828d2e3","source":{"kind":"arxiv","id":"2412.10146","version":2},"attestation_state":"computed","paper":{"title":"Investigating generalization capabilities of neural networks by means of loss landscapes and Hessian analysis","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Nikita Gabdullin","submitted_at":"2024-12-13T14:02:41Z","abstract_excerpt":"This paper studies generalization capabilities of neural networks (NNs) using new and improved PyTorch library Loss Landscape Analysis (LLA). LLA facilitates visualization and analysis of loss landscapes along with the properties of NN Hessian. Different approaches to NN loss landscape plotting are discussed with particular focus on normalization techniques showing that conventional methods cannot always ensure correct visualization when batch normalization layers are present in NN architecture. The use of Hessian axes is shown to be able to mitigate this effect, and methods for choosing Hessi"},"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":"2412.10146","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-12-13T14:02:41Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"779f33a87f86a4d3e67ce43f538e0bb7b24fdfe6df4a046c909be87c711a9cbf","abstract_canon_sha256":"85102f2de0cb22127b546dc22a87c360be441733f55924c4fa1216bb4b44bd25"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:09:50.601388Z","signature_b64":"4CVtdNOR8bWLsIn/NULWiZn+ntKhI4Gb93RxCxyei0KOvH4drG6CjY9wgsKNUECGmNCUVweXM4BOh7QKEBrGBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d3d55b75f2ee1029f5947d7049d72946ca60e9c6ae720ef76fc9b1db7828d2e3","last_reissued_at":"2026-07-05T10:09:50.600900Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:09:50.600900Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Investigating generalization capabilities of neural networks by means of loss landscapes and Hessian analysis","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Nikita Gabdullin","submitted_at":"2024-12-13T14:02:41Z","abstract_excerpt":"This paper studies generalization capabilities of neural networks (NNs) using new and improved PyTorch library Loss Landscape Analysis (LLA). LLA facilitates visualization and analysis of loss landscapes along with the properties of NN Hessian. Different approaches to NN loss landscape plotting are discussed with particular focus on normalization techniques showing that conventional methods cannot always ensure correct visualization when batch normalization layers are present in NN architecture. The use of Hessian axes is shown to be able to mitigate this effect, and methods for choosing Hessi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.10146","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/2412.10146/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":"2412.10146","created_at":"2026-07-05T10:09:50.600956+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.10146v2","created_at":"2026-07-05T10:09:50.600956+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.10146","created_at":"2026-07-05T10:09:50.600956+00:00"},{"alias_kind":"pith_short_12","alias_value":"2PKVW5PS5YIC","created_at":"2026-07-05T10:09:50.600956+00:00"},{"alias_kind":"pith_short_16","alias_value":"2PKVW5PS5YICT5MU","created_at":"2026-07-05T10:09:50.600956+00:00"},{"alias_kind":"pith_short_8","alias_value":"2PKVW5PS","created_at":"2026-07-05T10:09:50.600956+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2504.17618","citing_title":"The effects of Hessian eigenvalue spectral density type on the applicability of Hessian analysis to generalization capability assessment of neural networks","ref_index":5,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2PKVW5PS5YICT5MUPVYETVZJI3","json":"https://pith.science/pith/2PKVW5PS5YICT5MUPVYETVZJI3.json","graph_json":"https://pith.science/api/pith-number/2PKVW5PS5YICT5MUPVYETVZJI3/graph.json","events_json":"https://pith.science/api/pith-number/2PKVW5PS5YICT5MUPVYETVZJI3/events.json","paper":"https://pith.science/paper/2PKVW5PS"},"agent_actions":{"view_html":"https://pith.science/pith/2PKVW5PS5YICT5MUPVYETVZJI3","download_json":"https://pith.science/pith/2PKVW5PS5YICT5MUPVYETVZJI3.json","view_paper":"https://pith.science/paper/2PKVW5PS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.10146&json=true","fetch_graph":"https://pith.science/api/pith-number/2PKVW5PS5YICT5MUPVYETVZJI3/graph.json","fetch_events":"https://pith.science/api/pith-number/2PKVW5PS5YICT5MUPVYETVZJI3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2PKVW5PS5YICT5MUPVYETVZJI3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2PKVW5PS5YICT5MUPVYETVZJI3/action/storage_attestation","attest_author":"https://pith.science/pith/2PKVW5PS5YICT5MUPVYETVZJI3/action/author_attestation","sign_citation":"https://pith.science/pith/2PKVW5PS5YICT5MUPVYETVZJI3/action/citation_signature","submit_replication":"https://pith.science/pith/2PKVW5PS5YICT5MUPVYETVZJI3/action/replication_record"}},"created_at":"2026-07-05T10:09:50.600956+00:00","updated_at":"2026-07-05T10:09:50.600956+00:00"}