{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:37UT2RPVJ5JFVFCM4UKIQSTOY2","short_pith_number":"pith:37UT2RPV","schema_version":"1.0","canonical_sha256":"dfe93d45f54f525a944ce514884a6ec6aa2747850246df79622b8ed8e14855ce","source":{"kind":"arxiv","id":"2410.03030","version":2},"attestation_state":"computed","paper":{"title":"Dynamic Sparse Training versus Dense Training: The Unexpected Winner in Image Corruption Robustness","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Boqian Wu, Decebal Constantin Mocanu, Elena Mocanu, Maurice van Keulen, Mykola Pechenizkiy, Nicola Strisciuglio, Qiao Xiao, Shunxin Wang","submitted_at":"2024-10-03T22:24:54Z","abstract_excerpt":"It is generally perceived that Dynamic Sparse Training opens the door to a new era of scalability and efficiency for artificial neural networks at, perhaps, some costs in accuracy performance for the classification task. At the same time, Dense Training is widely accepted as being the \"de facto\" approach to train artificial neural networks if one would like to maximize their robustness against image corruption. In this paper, we question this general practice. Consequently, we claim that, contrary to what is commonly thought, the Dynamic Sparse Training methods can consistently outperform Dens"},"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":"2410.03030","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-10-03T22:24:54Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8969f68bda3b6e2081d86796e0d46589455b65ae52d017b29d74802991dc65e7","abstract_canon_sha256":"dfd4dc88ddb17934f8d3641fe14b35764449a91b7fdc0733dfdc11aac90cb567"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:24:26.235631Z","signature_b64":"Pq9FYsDS9zUzB3oBiFYSVBHDlW7/JE53+SL/WnmXzjYgOHCRDKxtHa9ZTvBG6a7qLex6FM/uPj28Awyo2zHPBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dfe93d45f54f525a944ce514884a6ec6aa2747850246df79622b8ed8e14855ce","last_reissued_at":"2026-07-05T10:24:26.234624Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:24:26.234624Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Dynamic Sparse Training versus Dense Training: The Unexpected Winner in Image Corruption Robustness","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Boqian Wu, Decebal Constantin Mocanu, Elena Mocanu, Maurice van Keulen, Mykola Pechenizkiy, Nicola Strisciuglio, Qiao Xiao, Shunxin Wang","submitted_at":"2024-10-03T22:24:54Z","abstract_excerpt":"It is generally perceived that Dynamic Sparse Training opens the door to a new era of scalability and efficiency for artificial neural networks at, perhaps, some costs in accuracy performance for the classification task. At the same time, Dense Training is widely accepted as being the \"de facto\" approach to train artificial neural networks if one would like to maximize their robustness against image corruption. In this paper, we question this general practice. Consequently, we claim that, contrary to what is commonly thought, the Dynamic Sparse Training methods can consistently outperform Dens"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.03030","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/2410.03030/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":"2410.03030","created_at":"2026-07-05T10:24:26.234764+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.03030v2","created_at":"2026-07-05T10:24:26.234764+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.03030","created_at":"2026-07-05T10:24:26.234764+00:00"},{"alias_kind":"pith_short_12","alias_value":"37UT2RPVJ5JF","created_at":"2026-07-05T10:24:26.234764+00:00"},{"alias_kind":"pith_short_16","alias_value":"37UT2RPVJ5JFVFCM","created_at":"2026-07-05T10:24:26.234764+00:00"},{"alias_kind":"pith_short_8","alias_value":"37UT2RPV","created_at":"2026-07-05T10:24:26.234764+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.17909","citing_title":"NeuroTrails: Training with Dynamic Sparse Heads as the Key to Effective Ensembling","ref_index":75,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/37UT2RPVJ5JFVFCM4UKIQSTOY2","json":"https://pith.science/pith/37UT2RPVJ5JFVFCM4UKIQSTOY2.json","graph_json":"https://pith.science/api/pith-number/37UT2RPVJ5JFVFCM4UKIQSTOY2/graph.json","events_json":"https://pith.science/api/pith-number/37UT2RPVJ5JFVFCM4UKIQSTOY2/events.json","paper":"https://pith.science/paper/37UT2RPV"},"agent_actions":{"view_html":"https://pith.science/pith/37UT2RPVJ5JFVFCM4UKIQSTOY2","download_json":"https://pith.science/pith/37UT2RPVJ5JFVFCM4UKIQSTOY2.json","view_paper":"https://pith.science/paper/37UT2RPV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.03030&json=true","fetch_graph":"https://pith.science/api/pith-number/37UT2RPVJ5JFVFCM4UKIQSTOY2/graph.json","fetch_events":"https://pith.science/api/pith-number/37UT2RPVJ5JFVFCM4UKIQSTOY2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/37UT2RPVJ5JFVFCM4UKIQSTOY2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/37UT2RPVJ5JFVFCM4UKIQSTOY2/action/storage_attestation","attest_author":"https://pith.science/pith/37UT2RPVJ5JFVFCM4UKIQSTOY2/action/author_attestation","sign_citation":"https://pith.science/pith/37UT2RPVJ5JFVFCM4UKIQSTOY2/action/citation_signature","submit_replication":"https://pith.science/pith/37UT2RPVJ5JFVFCM4UKIQSTOY2/action/replication_record"}},"created_at":"2026-07-05T10:24:26.234764+00:00","updated_at":"2026-07-05T10:24:26.234764+00:00"}