{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:BHESSHYMS6IUARA4OLT6NMJAGU","short_pith_number":"pith:BHESSHYM","schema_version":"1.0","canonical_sha256":"09c9291f0c979140441c72e7e6b120353e3e496404f374a0e98048b14ead938f","source":{"kind":"arxiv","id":"2302.03773","version":1},"attestation_state":"computed","paper":{"title":"What Matters In The Structured Pruning of Generative Language Models?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Michael Santacroce, Yelong Shen, Yuanzhi Li, Zixin Wen","submitted_at":"2023-02-07T22:05:55Z","abstract_excerpt":"Auto-regressive large language models such as GPT-3 require enormous computational resources to use. Traditionally, structured pruning methods are employed to reduce resource usage. However, their application to and efficacy for generative language models is heavily under-explored. In this paper we conduct an comprehensive evaluation of common structured pruning methods, including magnitude, random, and movement pruning on the feed-forward layers in GPT-type models. Unexpectedly, random pruning results in performance that is comparable to the best established methods, across multiple natural l"},"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":"2302.03773","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-02-07T22:05:55Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"ce2666fd913c10ecb53f5df302a1353b7b30036454365d6e9bba4e5138b013cd","abstract_canon_sha256":"d7748f5cd5b03bb2e8d008e6244074abf4c0219148d5860d6f2d9baa191caa7d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:39:57.503476Z","signature_b64":"SRSVQ+AMNObMLH16uEmHPfA7MPhP2kzyUbLrgE+lPT/EIvV9fNO5BVL/JFj9HjRtAvriVAoovM2d5+bOjL4uAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"09c9291f0c979140441c72e7e6b120353e3e496404f374a0e98048b14ead938f","last_reissued_at":"2026-07-05T05:39:57.502983Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:39:57.502983Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"What Matters In The Structured Pruning of Generative Language Models?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Michael Santacroce, Yelong Shen, Yuanzhi Li, Zixin Wen","submitted_at":"2023-02-07T22:05:55Z","abstract_excerpt":"Auto-regressive large language models such as GPT-3 require enormous computational resources to use. Traditionally, structured pruning methods are employed to reduce resource usage. However, their application to and efficacy for generative language models is heavily under-explored. In this paper we conduct an comprehensive evaluation of common structured pruning methods, including magnitude, random, and movement pruning on the feed-forward layers in GPT-type models. Unexpectedly, random pruning results in performance that is comparable to the best established methods, across multiple natural l"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2302.03773","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/2302.03773/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":"2302.03773","created_at":"2026-07-05T05:39:57.503045+00:00"},{"alias_kind":"arxiv_version","alias_value":"2302.03773v1","created_at":"2026-07-05T05:39:57.503045+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2302.03773","created_at":"2026-07-05T05:39:57.503045+00:00"},{"alias_kind":"pith_short_12","alias_value":"BHESSHYMS6IU","created_at":"2026-07-05T05:39:57.503045+00:00"},{"alias_kind":"pith_short_16","alias_value":"BHESSHYMS6IUARA4","created_at":"2026-07-05T05:39:57.503045+00:00"},{"alias_kind":"pith_short_8","alias_value":"BHESSHYM","created_at":"2026-07-05T05:39:57.503045+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2305.07759","citing_title":"TinyStories: How Small Can Language Models Be and Still Speak Coherent English?","ref_index":26,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BHESSHYMS6IUARA4OLT6NMJAGU","json":"https://pith.science/pith/BHESSHYMS6IUARA4OLT6NMJAGU.json","graph_json":"https://pith.science/api/pith-number/BHESSHYMS6IUARA4OLT6NMJAGU/graph.json","events_json":"https://pith.science/api/pith-number/BHESSHYMS6IUARA4OLT6NMJAGU/events.json","paper":"https://pith.science/paper/BHESSHYM"},"agent_actions":{"view_html":"https://pith.science/pith/BHESSHYMS6IUARA4OLT6NMJAGU","download_json":"https://pith.science/pith/BHESSHYMS6IUARA4OLT6NMJAGU.json","view_paper":"https://pith.science/paper/BHESSHYM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2302.03773&json=true","fetch_graph":"https://pith.science/api/pith-number/BHESSHYMS6IUARA4OLT6NMJAGU/graph.json","fetch_events":"https://pith.science/api/pith-number/BHESSHYMS6IUARA4OLT6NMJAGU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BHESSHYMS6IUARA4OLT6NMJAGU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BHESSHYMS6IUARA4OLT6NMJAGU/action/storage_attestation","attest_author":"https://pith.science/pith/BHESSHYMS6IUARA4OLT6NMJAGU/action/author_attestation","sign_citation":"https://pith.science/pith/BHESSHYMS6IUARA4OLT6NMJAGU/action/citation_signature","submit_replication":"https://pith.science/pith/BHESSHYMS6IUARA4OLT6NMJAGU/action/replication_record"}},"created_at":"2026-07-05T05:39:57.503045+00:00","updated_at":"2026-07-05T05:39:57.503045+00:00"}