{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:IXJQKLWGCJFYYUWWEKBUISD2NE","short_pith_number":"pith:IXJQKLWG","schema_version":"1.0","canonical_sha256":"45d3052ec6124b8c52d6228344487a692d9f3c8b16b7373489915987ce737aae","source":{"kind":"arxiv","id":"2210.03044","version":1},"attestation_state":"computed","paper":{"title":"Unmasking the Lottery Ticket Hypothesis: What's Encoded in a Winning Ticket's Mask?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Brett W. Larsen, Feng Chen, Gintare Karolina Dziugaite, Jonathan Frankle, Mansheej Paul, Surya Ganguli","submitted_at":"2022-10-06T16:50:20Z","abstract_excerpt":"Modern deep learning involves training costly, highly overparameterized networks, thus motivating the search for sparser networks that can still be trained to the same accuracy as the full network (i.e. matching). Iterative magnitude pruning (IMP) is a state of the art algorithm that can find such highly sparse matching subnetworks, known as winning tickets. IMP operates by iterative cycles of training, masking smallest magnitude weights, rewinding back to an early training point, and repeating. Despite its simplicity, the underlying principles for when and how IMP finds winning tickets remain"},"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":"2210.03044","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-10-06T16:50:20Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"4bfccc5d0ba17bd6537cbeeb77f6f2cdc2d51c4659e8a3fb832477eafabb217e","abstract_canon_sha256":"7981c15c91dc3d971eeaa9c4fd9be3068d8762da4e8d0cd73b7aaac0679e4865"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:04:06.656833Z","signature_b64":"ZyWOx3uG+yED8SPJdhWLvKBmCcdjSMUnK4+kSnhotaR4oRvDT/Ktqb1z6xAAWbXARX1kGEeeqc+lH3s9ris4Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"45d3052ec6124b8c52d6228344487a692d9f3c8b16b7373489915987ce737aae","last_reissued_at":"2026-07-05T05:04:06.656346Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:04:06.656346Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Unmasking the Lottery Ticket Hypothesis: What's Encoded in a Winning Ticket's Mask?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Brett W. Larsen, Feng Chen, Gintare Karolina Dziugaite, Jonathan Frankle, Mansheej Paul, Surya Ganguli","submitted_at":"2022-10-06T16:50:20Z","abstract_excerpt":"Modern deep learning involves training costly, highly overparameterized networks, thus motivating the search for sparser networks that can still be trained to the same accuracy as the full network (i.e. matching). Iterative magnitude pruning (IMP) is a state of the art algorithm that can find such highly sparse matching subnetworks, known as winning tickets. IMP operates by iterative cycles of training, masking smallest magnitude weights, rewinding back to an early training point, and repeating. Despite its simplicity, the underlying principles for when and how IMP finds winning tickets remain"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.03044","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/2210.03044/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":"2210.03044","created_at":"2026-07-05T05:04:06.656410+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.03044v1","created_at":"2026-07-05T05:04:06.656410+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.03044","created_at":"2026-07-05T05:04:06.656410+00:00"},{"alias_kind":"pith_short_12","alias_value":"IXJQKLWGCJFY","created_at":"2026-07-05T05:04:06.656410+00:00"},{"alias_kind":"pith_short_16","alias_value":"IXJQKLWGCJFYYUWW","created_at":"2026-07-05T05:04:06.656410+00:00"},{"alias_kind":"pith_short_8","alias_value":"IXJQKLWG","created_at":"2026-07-05T05:04:06.656410+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.19941","citing_title":"Compositionality Emerges in a Narrow Depth-Connectivity Regime: Architecture Constraints and Solution Manifolds","ref_index":71,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17704","citing_title":"Toy Combinatorial Interpretability Models Reveal Lottery Tickets in Early Feature Space","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2509.22166","citing_title":"Motivating Next-Gen Accelerators with Flexible (N:M) Activation Sparsity via Benchmarking Lightweight Post-Training Sparsification Approaches","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IXJQKLWGCJFYYUWWEKBUISD2NE","json":"https://pith.science/pith/IXJQKLWGCJFYYUWWEKBUISD2NE.json","graph_json":"https://pith.science/api/pith-number/IXJQKLWGCJFYYUWWEKBUISD2NE/graph.json","events_json":"https://pith.science/api/pith-number/IXJQKLWGCJFYYUWWEKBUISD2NE/events.json","paper":"https://pith.science/paper/IXJQKLWG"},"agent_actions":{"view_html":"https://pith.science/pith/IXJQKLWGCJFYYUWWEKBUISD2NE","download_json":"https://pith.science/pith/IXJQKLWGCJFYYUWWEKBUISD2NE.json","view_paper":"https://pith.science/paper/IXJQKLWG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.03044&json=true","fetch_graph":"https://pith.science/api/pith-number/IXJQKLWGCJFYYUWWEKBUISD2NE/graph.json","fetch_events":"https://pith.science/api/pith-number/IXJQKLWGCJFYYUWWEKBUISD2NE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IXJQKLWGCJFYYUWWEKBUISD2NE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IXJQKLWGCJFYYUWWEKBUISD2NE/action/storage_attestation","attest_author":"https://pith.science/pith/IXJQKLWGCJFYYUWWEKBUISD2NE/action/author_attestation","sign_citation":"https://pith.science/pith/IXJQKLWGCJFYYUWWEKBUISD2NE/action/citation_signature","submit_replication":"https://pith.science/pith/IXJQKLWGCJFYYUWWEKBUISD2NE/action/replication_record"}},"created_at":"2026-07-05T05:04:06.656410+00:00","updated_at":"2026-07-05T05:04:06.656410+00:00"}