{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:DMF2UFQ2SZUKISPMD7AI2E7MSO","short_pith_number":"pith:DMF2UFQ2","schema_version":"1.0","canonical_sha256":"1b0baa161a9668a449ec1fc08d13ec9384a45af770e04e0f0e47f4e4320ea12e","source":{"kind":"arxiv","id":"2106.01429","version":2},"attestation_state":"computed","paper":{"title":"Smooth Bilevel Programming for Sparse Regularization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","math.OC"],"primary_cat":"stat.ML","authors_text":"Clarice Poon, Gabriel Peyr\\'e","submitted_at":"2021-06-02T19:18:22Z","abstract_excerpt":"Iteratively reweighted least square (IRLS) is a popular approach to solve sparsity-enforcing regression problems in machine learning. State of the art approaches are more efficient but typically rely on specific coordinate pruning schemes. In this work, we show how a surprisingly simple reparametrization of IRLS, coupled with a bilevel resolution (instead of an alternating scheme) is able to achieve top performances on a wide range of sparsity (such as Lasso, group Lasso and trace norm regularizations), regularization strength (including hard constraints), and design matrices (ranging from cor"},"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":"2106.01429","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2021-06-02T19:18:22Z","cross_cats_sorted":["cs.LG","math.OC"],"title_canon_sha256":"87fcd20ce403bc2ea21f41cf974e8d69cdeda38cac4656eb07f50b8e66e697a2","abstract_canon_sha256":"37b4648c551152b410a28e2828ae672ed7e1bed4e786b5b95fd279071f4abb83"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:02:08.918729Z","signature_b64":"TJ0RXpdYRceHb5rzUry+i95Lu9bBoPrzgck/Uj6Wsolo0UnDqUaxRhKRK06FMIJPN8ql7vkj4k4949q5NZaYDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1b0baa161a9668a449ec1fc08d13ec9384a45af770e04e0f0e47f4e4320ea12e","last_reissued_at":"2026-07-05T05:02:08.918294Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:02:08.918294Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Smooth Bilevel Programming for Sparse Regularization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","math.OC"],"primary_cat":"stat.ML","authors_text":"Clarice Poon, Gabriel Peyr\\'e","submitted_at":"2021-06-02T19:18:22Z","abstract_excerpt":"Iteratively reweighted least square (IRLS) is a popular approach to solve sparsity-enforcing regression problems in machine learning. State of the art approaches are more efficient but typically rely on specific coordinate pruning schemes. In this work, we show how a surprisingly simple reparametrization of IRLS, coupled with a bilevel resolution (instead of an alternating scheme) is able to achieve top performances on a wide range of sparsity (such as Lasso, group Lasso and trace norm regularizations), regularization strength (including hard constraints), and design matrices (ranging from cor"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.01429","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/2106.01429/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":"2106.01429","created_at":"2026-07-05T05:02:08.918358+00:00"},{"alias_kind":"arxiv_version","alias_value":"2106.01429v2","created_at":"2026-07-05T05:02:08.918358+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.01429","created_at":"2026-07-05T05:02:08.918358+00:00"},{"alias_kind":"pith_short_12","alias_value":"DMF2UFQ2SZUK","created_at":"2026-07-05T05:02:08.918358+00:00"},{"alias_kind":"pith_short_16","alias_value":"DMF2UFQ2SZUKISPM","created_at":"2026-07-05T05:02:08.918358+00:00"},{"alias_kind":"pith_short_8","alias_value":"DMF2UFQ2","created_at":"2026-07-05T05:02:08.918358+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.00013","citing_title":"From Provable Correctness to Probabilistic Generation: A Comparative Review of Program Synthesis Paradigms","ref_index":12,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DMF2UFQ2SZUKISPMD7AI2E7MSO","json":"https://pith.science/pith/DMF2UFQ2SZUKISPMD7AI2E7MSO.json","graph_json":"https://pith.science/api/pith-number/DMF2UFQ2SZUKISPMD7AI2E7MSO/graph.json","events_json":"https://pith.science/api/pith-number/DMF2UFQ2SZUKISPMD7AI2E7MSO/events.json","paper":"https://pith.science/paper/DMF2UFQ2"},"agent_actions":{"view_html":"https://pith.science/pith/DMF2UFQ2SZUKISPMD7AI2E7MSO","download_json":"https://pith.science/pith/DMF2UFQ2SZUKISPMD7AI2E7MSO.json","view_paper":"https://pith.science/paper/DMF2UFQ2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2106.01429&json=true","fetch_graph":"https://pith.science/api/pith-number/DMF2UFQ2SZUKISPMD7AI2E7MSO/graph.json","fetch_events":"https://pith.science/api/pith-number/DMF2UFQ2SZUKISPMD7AI2E7MSO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DMF2UFQ2SZUKISPMD7AI2E7MSO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DMF2UFQ2SZUKISPMD7AI2E7MSO/action/storage_attestation","attest_author":"https://pith.science/pith/DMF2UFQ2SZUKISPMD7AI2E7MSO/action/author_attestation","sign_citation":"https://pith.science/pith/DMF2UFQ2SZUKISPMD7AI2E7MSO/action/citation_signature","submit_replication":"https://pith.science/pith/DMF2UFQ2SZUKISPMD7AI2E7MSO/action/replication_record"}},"created_at":"2026-07-05T05:02:08.918358+00:00","updated_at":"2026-07-05T05:02:08.918358+00:00"}