{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:LTODQT4L2TRDGSMGBIWOG4WH2A","short_pith_number":"pith:LTODQT4L","schema_version":"1.0","canonical_sha256":"5cdc384f8bd4e23349860a2ce372c7d01b4f1fd2350a5f410fc1631efd4f3596","source":{"kind":"arxiv","id":"2409.15156","version":2},"attestation_state":"computed","paper":{"title":"Rethinking Conventional Wisdom in Machine Learning: From Generalization to Scaling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Lechao Xiao","submitted_at":"2024-09-23T16:04:03Z","abstract_excerpt":"The remarkable success of large language pretraining and the discovery of scaling laws signify a paradigm shift in machine learning. Notably, the primary objective has evolved from minimizing generalization error to reducing approximation error, and the most effective strategy has transitioned from regularization (in a broad sense) to scaling up models. This raises a critical question:\n  Do the established principles that proved successful in the generalization-centric era remain valid in this new era of scaling?\n  This paper examines several influential regularization-based principles that ma"},"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":"2409.15156","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-09-23T16:04:03Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"36585729df8e67d93f12d4b89c36be5866895b88bc61f8f3e7f1cea2966a14a8","abstract_canon_sha256":"150a1f52925c88ba271798e48cd11d28cb9ac493a23d2c788aec7f4e38fbfa06"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:14:05.612941Z","signature_b64":"fYnex53lzHXJ5t/AVgoRbHaSTOxxGQDnFrhfzOnwK2byl6VZQst1L/MKVEgqfFD8x+VUKuwu9vfVZ1bMV9O3BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5cdc384f8bd4e23349860a2ce372c7d01b4f1fd2350a5f410fc1631efd4f3596","last_reissued_at":"2026-07-05T10:14:05.612447Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:14:05.612447Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rethinking Conventional Wisdom in Machine Learning: From Generalization to Scaling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Lechao Xiao","submitted_at":"2024-09-23T16:04:03Z","abstract_excerpt":"The remarkable success of large language pretraining and the discovery of scaling laws signify a paradigm shift in machine learning. Notably, the primary objective has evolved from minimizing generalization error to reducing approximation error, and the most effective strategy has transitioned from regularization (in a broad sense) to scaling up models. This raises a critical question:\n  Do the established principles that proved successful in the generalization-centric era remain valid in this new era of scaling?\n  This paper examines several influential regularization-based principles that ma"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.15156","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/2409.15156/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":"2409.15156","created_at":"2026-07-05T10:14:05.612506+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.15156v2","created_at":"2026-07-05T10:14:05.612506+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.15156","created_at":"2026-07-05T10:14:05.612506+00:00"},{"alias_kind":"pith_short_12","alias_value":"LTODQT4L2TRD","created_at":"2026-07-05T10:14:05.612506+00:00"},{"alias_kind":"pith_short_16","alias_value":"LTODQT4L2TRDGSMG","created_at":"2026-07-05T10:14:05.612506+00:00"},{"alias_kind":"pith_short_8","alias_value":"LTODQT4L","created_at":"2026-07-05T10:14:05.612506+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.18528","citing_title":"Scale-Invariant Neural Network Optimization: Norm Geometry and Heavy-Tailed Noise","ref_index":124,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29303","citing_title":"Entropy-KL Divergence-based Token Masking: A Novel Approach for Selective Fine-tuning of Large Language Models","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18528","citing_title":"Scale-Invariant Neural Network Optimization: Norm Geometry and Heavy-Tailed Noise","ref_index":133,"is_internal_anchor":false},{"citing_arxiv_id":"2508.17784","citing_title":"Proximal Supervised Fine-Tuning","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11172","citing_title":"Optimistic Dual Averaging Unifies Modern Optimizers","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20244","citing_title":"Hybrid Policy Distillation for LLMs","ref_index":34,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LTODQT4L2TRDGSMGBIWOG4WH2A","json":"https://pith.science/pith/LTODQT4L2TRDGSMGBIWOG4WH2A.json","graph_json":"https://pith.science/api/pith-number/LTODQT4L2TRDGSMGBIWOG4WH2A/graph.json","events_json":"https://pith.science/api/pith-number/LTODQT4L2TRDGSMGBIWOG4WH2A/events.json","paper":"https://pith.science/paper/LTODQT4L"},"agent_actions":{"view_html":"https://pith.science/pith/LTODQT4L2TRDGSMGBIWOG4WH2A","download_json":"https://pith.science/pith/LTODQT4L2TRDGSMGBIWOG4WH2A.json","view_paper":"https://pith.science/paper/LTODQT4L","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.15156&json=true","fetch_graph":"https://pith.science/api/pith-number/LTODQT4L2TRDGSMGBIWOG4WH2A/graph.json","fetch_events":"https://pith.science/api/pith-number/LTODQT4L2TRDGSMGBIWOG4WH2A/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LTODQT4L2TRDGSMGBIWOG4WH2A/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LTODQT4L2TRDGSMGBIWOG4WH2A/action/storage_attestation","attest_author":"https://pith.science/pith/LTODQT4L2TRDGSMGBIWOG4WH2A/action/author_attestation","sign_citation":"https://pith.science/pith/LTODQT4L2TRDGSMGBIWOG4WH2A/action/citation_signature","submit_replication":"https://pith.science/pith/LTODQT4L2TRDGSMGBIWOG4WH2A/action/replication_record"}},"created_at":"2026-07-05T10:14:05.612506+00:00","updated_at":"2026-07-05T10:14:05.612506+00:00"}