{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:WS6XY5MLBKNTE25NGJ62IOVYFU","short_pith_number":"pith:WS6XY5ML","schema_version":"1.0","canonical_sha256":"b4bd7c758b0a9b326bad327da43ab82d2300d08e906053e776eb20853b8eec22","source":{"kind":"arxiv","id":"2203.02714","version":1},"attestation_state":"computed","paper":{"title":"Towards Efficient and Scalable Sharpness-Aware Minimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Cho-Jui Hsieh, Siqi Mai, Xiangning Chen, Yang You, Yong Liu","submitted_at":"2022-03-05T11:53:37Z","abstract_excerpt":"Recently, Sharpness-Aware Minimization (SAM), which connects the geometry of the loss landscape and generalization, has demonstrated significant performance boosts on training large-scale models such as vision transformers. However, the update rule of SAM requires two sequential (non-parallelizable) gradient computations at each step, which can double the computational overhead. In this paper, we propose a novel algorithm LookSAM - that only periodically calculates the inner gradient ascent, to significantly reduce the additional training cost of SAM. The empirical results illustrate that Look"},"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":"2203.02714","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-03-05T11:53:37Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"7d243f525769fc8356accf490fa86ceeafd643d2efe5fbcdec9de7560efa2346","abstract_canon_sha256":"5b12e6d4e3cf5f2d5ca8a261be9c5979d2b4b28c2096f8a6f1f7f21bf08057ff"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:02:26.434134Z","signature_b64":"2+ccFTQAmWKOe0vc0LDn0Zl9klu/f/w1UM60o0YOltCs6RhUaz1ytIcNvwxEnCHIyCySB0mPAtuBACPYN8EABg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b4bd7c758b0a9b326bad327da43ab82d2300d08e906053e776eb20853b8eec22","last_reissued_at":"2026-07-05T04:02:26.433652Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:02:26.433652Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards Efficient and Scalable Sharpness-Aware Minimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Cho-Jui Hsieh, Siqi Mai, Xiangning Chen, Yang You, Yong Liu","submitted_at":"2022-03-05T11:53:37Z","abstract_excerpt":"Recently, Sharpness-Aware Minimization (SAM), which connects the geometry of the loss landscape and generalization, has demonstrated significant performance boosts on training large-scale models such as vision transformers. However, the update rule of SAM requires two sequential (non-parallelizable) gradient computations at each step, which can double the computational overhead. In this paper, we propose a novel algorithm LookSAM - that only periodically calculates the inner gradient ascent, to significantly reduce the additional training cost of SAM. The empirical results illustrate that Look"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2203.02714","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/2203.02714/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":"2203.02714","created_at":"2026-07-05T04:02:26.433711+00:00"},{"alias_kind":"arxiv_version","alias_value":"2203.02714v1","created_at":"2026-07-05T04:02:26.433711+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2203.02714","created_at":"2026-07-05T04:02:26.433711+00:00"},{"alias_kind":"pith_short_12","alias_value":"WS6XY5MLBKNT","created_at":"2026-07-05T04:02:26.433711+00:00"},{"alias_kind":"pith_short_16","alias_value":"WS6XY5MLBKNTE25N","created_at":"2026-07-05T04:02:26.433711+00:00"},{"alias_kind":"pith_short_8","alias_value":"WS6XY5ML","created_at":"2026-07-05T04:02:26.433711+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2509.03110","citing_title":"LSAM: Asynchronous Distributed Training with Landscape-Smoothed Sharpness-Aware Minimization","ref_index":35,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WS6XY5MLBKNTE25NGJ62IOVYFU","json":"https://pith.science/pith/WS6XY5MLBKNTE25NGJ62IOVYFU.json","graph_json":"https://pith.science/api/pith-number/WS6XY5MLBKNTE25NGJ62IOVYFU/graph.json","events_json":"https://pith.science/api/pith-number/WS6XY5MLBKNTE25NGJ62IOVYFU/events.json","paper":"https://pith.science/paper/WS6XY5ML"},"agent_actions":{"view_html":"https://pith.science/pith/WS6XY5MLBKNTE25NGJ62IOVYFU","download_json":"https://pith.science/pith/WS6XY5MLBKNTE25NGJ62IOVYFU.json","view_paper":"https://pith.science/paper/WS6XY5ML","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2203.02714&json=true","fetch_graph":"https://pith.science/api/pith-number/WS6XY5MLBKNTE25NGJ62IOVYFU/graph.json","fetch_events":"https://pith.science/api/pith-number/WS6XY5MLBKNTE25NGJ62IOVYFU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WS6XY5MLBKNTE25NGJ62IOVYFU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WS6XY5MLBKNTE25NGJ62IOVYFU/action/storage_attestation","attest_author":"https://pith.science/pith/WS6XY5MLBKNTE25NGJ62IOVYFU/action/author_attestation","sign_citation":"https://pith.science/pith/WS6XY5MLBKNTE25NGJ62IOVYFU/action/citation_signature","submit_replication":"https://pith.science/pith/WS6XY5MLBKNTE25NGJ62IOVYFU/action/replication_record"}},"created_at":"2026-07-05T04:02:26.433711+00:00","updated_at":"2026-07-05T04:02:26.433711+00:00"}