{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:LZL23LLCNDUULLDJWBTFFYM5UR","short_pith_number":"pith:LZL23LLC","schema_version":"1.0","canonical_sha256":"5e57adad6268e945ac69b06652e19da4687a07c08e9ba7e4835e2b8ccc12da3e","source":{"kind":"arxiv","id":"2511.14282","version":2},"attestation_state":"computed","paper":{"title":"Weight Concentration Regularization for Improving Pruning Robustness Under High Sparsity","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Junhyuk Jo, Sunwoo Lee, Vincent-Daniel Yun","submitted_at":"2025-11-18T09:18:26Z","abstract_excerpt":"Deep neural networks achieve outstanding performance across vision and language tasks, yet their large parameter counts limit deployment in resource-constrained settings. One-shot pruning reduces model size without retraining, but models trained with standard objectives often suffer substantial accuracy drops under aggressive sparsity. Prior work mitigates this drop along two directions: regularizers such as $\\ell_1$ and DeepHoyer that shape the weight distribution during training, and pruning-robust optimizers such as SAM, CrAM, and S$^2$SAM that flatten the loss landscape. However, existing "},"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":"2511.14282","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-18T09:18:26Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"676f6d00679a5630ccbeafad05558abb45e45ab5d4aa256f389c3ce26eba8365","abstract_canon_sha256":"665d27526ba2a0b65141d598b5896528ad9d2f28b3e2558034a0d632c3f910e4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:00:28.050160Z","signature_b64":"ZAXC1CmG+nJWCWzuxRnxxc4kg+0syZFb7SWdQEAMTbRJ14pL4BupRacqODuvIPHnVGDc37DQXPhURIfmUcbjDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5e57adad6268e945ac69b06652e19da4687a07c08e9ba7e4835e2b8ccc12da3e","last_reissued_at":"2026-05-20T00:00:28.049348Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:00:28.049348Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Weight Concentration Regularization for Improving Pruning Robustness Under High Sparsity","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Junhyuk Jo, Sunwoo Lee, Vincent-Daniel Yun","submitted_at":"2025-11-18T09:18:26Z","abstract_excerpt":"Deep neural networks achieve outstanding performance across vision and language tasks, yet their large parameter counts limit deployment in resource-constrained settings. One-shot pruning reduces model size without retraining, but models trained with standard objectives often suffer substantial accuracy drops under aggressive sparsity. Prior work mitigates this drop along two directions: regularizers such as $\\ell_1$ and DeepHoyer that shape the weight distribution during training, and pruning-robust optimizers such as SAM, CrAM, and S$^2$SAM that flatten the loss landscape. However, existing "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2511.14282","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/2511.14282/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":"2511.14282","created_at":"2026-05-20T00:00:28.049504+00:00"},{"alias_kind":"arxiv_version","alias_value":"2511.14282v2","created_at":"2026-05-20T00:00:28.049504+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.14282","created_at":"2026-05-20T00:00:28.049504+00:00"},{"alias_kind":"pith_short_12","alias_value":"LZL23LLCNDUU","created_at":"2026-05-20T00:00:28.049504+00:00"},{"alias_kind":"pith_short_16","alias_value":"LZL23LLCNDUULLDJ","created_at":"2026-05-20T00:00:28.049504+00:00"},{"alias_kind":"pith_short_8","alias_value":"LZL23LLC","created_at":"2026-05-20T00:00:28.049504+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LZL23LLCNDUULLDJWBTFFYM5UR","json":"https://pith.science/pith/LZL23LLCNDUULLDJWBTFFYM5UR.json","graph_json":"https://pith.science/api/pith-number/LZL23LLCNDUULLDJWBTFFYM5UR/graph.json","events_json":"https://pith.science/api/pith-number/LZL23LLCNDUULLDJWBTFFYM5UR/events.json","paper":"https://pith.science/paper/LZL23LLC"},"agent_actions":{"view_html":"https://pith.science/pith/LZL23LLCNDUULLDJWBTFFYM5UR","download_json":"https://pith.science/pith/LZL23LLCNDUULLDJWBTFFYM5UR.json","view_paper":"https://pith.science/paper/LZL23LLC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2511.14282&json=true","fetch_graph":"https://pith.science/api/pith-number/LZL23LLCNDUULLDJWBTFFYM5UR/graph.json","fetch_events":"https://pith.science/api/pith-number/LZL23LLCNDUULLDJWBTFFYM5UR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LZL23LLCNDUULLDJWBTFFYM5UR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LZL23LLCNDUULLDJWBTFFYM5UR/action/storage_attestation","attest_author":"https://pith.science/pith/LZL23LLCNDUULLDJWBTFFYM5UR/action/author_attestation","sign_citation":"https://pith.science/pith/LZL23LLCNDUULLDJWBTFFYM5UR/action/citation_signature","submit_replication":"https://pith.science/pith/LZL23LLCNDUULLDJWBTFFYM5UR/action/replication_record"}},"created_at":"2026-05-20T00:00:28.049504+00:00","updated_at":"2026-05-20T00:00:28.049504+00:00"}