{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:Q6DSLZUZMSHWPUADGCYVLHIGWY","short_pith_number":"pith:Q6DSLZUZ","schema_version":"1.0","canonical_sha256":"878725e699648f67d00330b1559d06b60bde97a75d092f12a6a633210abf339d","source":{"kind":"arxiv","id":"2403.10459","version":1},"attestation_state":"computed","paper":{"title":"Understanding the Double Descent Phenomenon in Deep Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Alexandre Thomas, Marc Lafon","submitted_at":"2024-03-15T16:51:24Z","abstract_excerpt":"Combining empirical risk minimization with capacity control is a classical strategy in machine learning when trying to control the generalization gap and avoid overfitting, as the model class capacity gets larger. Yet, in modern deep learning practice, very large over-parameterized models (e.g. neural networks) are optimized to fit perfectly the training data and still obtain great generalization performance. Past the interpolation point, increasing model complexity seems to actually lower the test error.\n  In this tutorial, we explain the concept of double descent and its mechanisms. The firs"},"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":"2403.10459","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-03-15T16:51:24Z","cross_cats_sorted":["cs.CV","stat.ML"],"title_canon_sha256":"16f2a2ceec7e30327d68f414864c201df20fca8dd012d563c5a5959e54b2940d","abstract_canon_sha256":"8d59fbe28b7b90965dcd0b8fd65fa4d6b59a5d613f2511651afb3bc97059de91"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:56:41.754807Z","signature_b64":"JVwiVSA02YVE6V8WFiBACtKvFea389JQDJP2faYZ0gnorUD4kg3+hWYbXAPeO5agc0LekPziC7G8tROtGGzFDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"878725e699648f67d00330b1559d06b60bde97a75d092f12a6a633210abf339d","last_reissued_at":"2026-07-05T07:56:41.754276Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:56:41.754276Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Understanding the Double Descent Phenomenon in Deep Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Alexandre Thomas, Marc Lafon","submitted_at":"2024-03-15T16:51:24Z","abstract_excerpt":"Combining empirical risk minimization with capacity control is a classical strategy in machine learning when trying to control the generalization gap and avoid overfitting, as the model class capacity gets larger. Yet, in modern deep learning practice, very large over-parameterized models (e.g. neural networks) are optimized to fit perfectly the training data and still obtain great generalization performance. Past the interpolation point, increasing model complexity seems to actually lower the test error.\n  In this tutorial, we explain the concept of double descent and its mechanisms. The firs"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.10459","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/2403.10459/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":"2403.10459","created_at":"2026-07-05T07:56:41.754344+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.10459v1","created_at":"2026-07-05T07:56:41.754344+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.10459","created_at":"2026-07-05T07:56:41.754344+00:00"},{"alias_kind":"pith_short_12","alias_value":"Q6DSLZUZMSHW","created_at":"2026-07-05T07:56:41.754344+00:00"},{"alias_kind":"pith_short_16","alias_value":"Q6DSLZUZMSHWPUAD","created_at":"2026-07-05T07:56:41.754344+00:00"},{"alias_kind":"pith_short_8","alias_value":"Q6DSLZUZ","created_at":"2026-07-05T07:56:41.754344+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2512.08444","citing_title":"Learned iterative networks: An operator learning perspective","ref_index":95,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/Q6DSLZUZMSHWPUADGCYVLHIGWY","json":"https://pith.science/pith/Q6DSLZUZMSHWPUADGCYVLHIGWY.json","graph_json":"https://pith.science/api/pith-number/Q6DSLZUZMSHWPUADGCYVLHIGWY/graph.json","events_json":"https://pith.science/api/pith-number/Q6DSLZUZMSHWPUADGCYVLHIGWY/events.json","paper":"https://pith.science/paper/Q6DSLZUZ"},"agent_actions":{"view_html":"https://pith.science/pith/Q6DSLZUZMSHWPUADGCYVLHIGWY","download_json":"https://pith.science/pith/Q6DSLZUZMSHWPUADGCYVLHIGWY.json","view_paper":"https://pith.science/paper/Q6DSLZUZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.10459&json=true","fetch_graph":"https://pith.science/api/pith-number/Q6DSLZUZMSHWPUADGCYVLHIGWY/graph.json","fetch_events":"https://pith.science/api/pith-number/Q6DSLZUZMSHWPUADGCYVLHIGWY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Q6DSLZUZMSHWPUADGCYVLHIGWY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Q6DSLZUZMSHWPUADGCYVLHIGWY/action/storage_attestation","attest_author":"https://pith.science/pith/Q6DSLZUZMSHWPUADGCYVLHIGWY/action/author_attestation","sign_citation":"https://pith.science/pith/Q6DSLZUZMSHWPUADGCYVLHIGWY/action/citation_signature","submit_replication":"https://pith.science/pith/Q6DSLZUZMSHWPUADGCYVLHIGWY/action/replication_record"}},"created_at":"2026-07-05T07:56:41.754344+00:00","updated_at":"2026-07-05T07:56:41.754344+00:00"}