{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:52QMVF6BIVA3GFSBUQSQZX37TT","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"ebca7244159be0fc25d41611e366d1cf976867491eda13351a8a96c42ad77dc9","cross_cats_sorted":["cs.DS","cs.NE","math.OC","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-01-13T17:28:29Z","title_canon_sha256":"88ec60b3cb02e31cd93dad67794af9e3cbe80a94e14b02ab58be16a65183158e"},"schema_version":"1.0","source":{"id":"2001.04413","kind":"arxiv","version":6}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2001.04413","created_at":"2026-07-05T06:28:29Z"},{"alias_kind":"arxiv_version","alias_value":"2001.04413v6","created_at":"2026-07-05T06:28:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2001.04413","created_at":"2026-07-05T06:28:29Z"},{"alias_kind":"pith_short_12","alias_value":"52QMVF6BIVA3","created_at":"2026-07-05T06:28:29Z"},{"alias_kind":"pith_short_16","alias_value":"52QMVF6BIVA3GFSB","created_at":"2026-07-05T06:28:29Z"},{"alias_kind":"pith_short_8","alias_value":"52QMVF6B","created_at":"2026-07-05T06:28:29Z"}],"graph_snapshots":[{"event_id":"sha256:c4f67151a6073a87568e28656218bc982a0c270cacbb52ec3ab2228094f109ee","target":"graph","created_at":"2026-07-05T06:28:29Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2001.04413/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Deep learning is also known as hierarchical learning, where the learner _learns_ to represent a complicated target function by decomposing it into a sequence of simpler functions to reduce sample and time complexity. This paper formally analyzes how multi-layer neural networks can perform such hierarchical learning _efficiently_ and _automatically_ by SGD on the training objective.\n  On the conceptual side, we present a theoretical characterizations of how certain types of deep (i.e. super-constant layer) neural networks can still be sample and time efficiently trained on some hierarchical tas","authors_text":"Yuanzhi Li, Zeyuan Allen-Zhu","cross_cats":["cs.DS","cs.NE","math.OC","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-01-13T17:28:29Z","title":"Backward Feature Correction: How Deep Learning Performs Deep (Hierarchical) Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2001.04413","kind":"arxiv","version":6},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:43e63b7aa35fc5ecaa3615388ab8546886d27eeaad80ed80a997dd249628d054","target":"record","created_at":"2026-07-05T06:28:29Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"ebca7244159be0fc25d41611e366d1cf976867491eda13351a8a96c42ad77dc9","cross_cats_sorted":["cs.DS","cs.NE","math.OC","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-01-13T17:28:29Z","title_canon_sha256":"88ec60b3cb02e31cd93dad67794af9e3cbe80a94e14b02ab58be16a65183158e"},"schema_version":"1.0","source":{"id":"2001.04413","kind":"arxiv","version":6}},"canonical_sha256":"eea0ca97c14541b31641a4250cdf7f9cd1f07c8e59897916ff291c9576e5c360","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"eea0ca97c14541b31641a4250cdf7f9cd1f07c8e59897916ff291c9576e5c360","first_computed_at":"2026-07-05T06:28:29.249053Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T06:28:29.249053Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"aV+NjCcc795/5+en3MPScbqqx9atZ+Wi2aGrEeFoZxprP3N85WS7ODmOsa8oQDwVmJiq+DgVbKFwyz4JTIDKDg==","signature_status":"signed_v1","signed_at":"2026-07-05T06:28:29.249523Z","signed_message":"canonical_sha256_bytes"},"source_id":"2001.04413","source_kind":"arxiv","source_version":6}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:43e63b7aa35fc5ecaa3615388ab8546886d27eeaad80ed80a997dd249628d054","sha256:c4f67151a6073a87568e28656218bc982a0c270cacbb52ec3ab2228094f109ee"],"state_sha256":"73d0e9f4a557625a3cf7794f04e9b7300d0fad16528e44416e0808f0021b82b1"}