{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:IX2K6MDJWBAL333S5UPOJE4TN5","short_pith_number":"pith:IX2K6MDJ","schema_version":"1.0","canonical_sha256":"45f4af3069b040bdef72ed1ee493936f67348ab57b28dc42455bbc5f78c4442a","source":{"kind":"arxiv","id":"2401.16405","version":2},"attestation_state":"computed","paper":{"title":"Scaling Sparse Fine-Tuning to Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Alan Ansell, Anna Korhonen, Edoardo M. Ponti, Hannah Sterz, Ivan Vuli\\'c","submitted_at":"2024-01-29T18:43:49Z","abstract_excerpt":"Large Language Models (LLMs) are difficult to fully fine-tune (e.g., with instructions or human feedback) due to their sheer number of parameters. A family of parameter-efficient sparse fine-tuning methods have proven promising in terms of performance but their memory requirements increase proportionally to the size of the LLMs. In this work, we scale sparse fine-tuning to state-of-the-art LLMs like LLaMA 2 7B and 13B. We propose SpIEL, a novel sparse fine-tuning method which, for a desired density level, maintains an array of parameter indices and the deltas of these parameters relative to th"},"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":"2401.16405","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-01-29T18:43:49Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"ac6b6fb08c9559a9a9a1583a3753e703a4198f19a1eca48052f07e97e8e99211","abstract_canon_sha256":"d40b0b4259a2e331f9a36f1a5e6b0cff04c938227611b3183c47063c634eff26"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:40:43.711267Z","signature_b64":"f17Y1qwy7jC/sT4gHLSYtH2TEGNu2uU4n2nilyg1hZKEtRRPsKeeDTcicVp99jexx9LC6/leATfQTKzidFGUBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"45f4af3069b040bdef72ed1ee493936f67348ab57b28dc42455bbc5f78c4442a","last_reissued_at":"2026-07-05T07:40:43.710690Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:40:43.710690Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Scaling Sparse Fine-Tuning to Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Alan Ansell, Anna Korhonen, Edoardo M. Ponti, Hannah Sterz, Ivan Vuli\\'c","submitted_at":"2024-01-29T18:43:49Z","abstract_excerpt":"Large Language Models (LLMs) are difficult to fully fine-tune (e.g., with instructions or human feedback) due to their sheer number of parameters. A family of parameter-efficient sparse fine-tuning methods have proven promising in terms of performance but their memory requirements increase proportionally to the size of the LLMs. In this work, we scale sparse fine-tuning to state-of-the-art LLMs like LLaMA 2 7B and 13B. We propose SpIEL, a novel sparse fine-tuning method which, for a desired density level, maintains an array of parameter indices and the deltas of these parameters relative to th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.16405","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/2401.16405/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":"2401.16405","created_at":"2026-07-05T07:40:43.710750+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.16405v2","created_at":"2026-07-05T07:40:43.710750+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.16405","created_at":"2026-07-05T07:40:43.710750+00:00"},{"alias_kind":"pith_short_12","alias_value":"IX2K6MDJWBAL","created_at":"2026-07-05T07:40:43.710750+00:00"},{"alias_kind":"pith_short_16","alias_value":"IX2K6MDJWBAL333S","created_at":"2026-07-05T07:40:43.710750+00:00"},{"alias_kind":"pith_short_8","alias_value":"IX2K6MDJ","created_at":"2026-07-05T07:40:43.710750+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.01155","citing_title":"When Data Is Scarce: Scaling Sparse Language Models with Repeated Training","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06166","citing_title":"One Algorithm, Two Goals: Dual Scoring for Parameter and Data Selection in LLM Fine-Tuning","ref_index":49,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IX2K6MDJWBAL333S5UPOJE4TN5","json":"https://pith.science/pith/IX2K6MDJWBAL333S5UPOJE4TN5.json","graph_json":"https://pith.science/api/pith-number/IX2K6MDJWBAL333S5UPOJE4TN5/graph.json","events_json":"https://pith.science/api/pith-number/IX2K6MDJWBAL333S5UPOJE4TN5/events.json","paper":"https://pith.science/paper/IX2K6MDJ"},"agent_actions":{"view_html":"https://pith.science/pith/IX2K6MDJWBAL333S5UPOJE4TN5","download_json":"https://pith.science/pith/IX2K6MDJWBAL333S5UPOJE4TN5.json","view_paper":"https://pith.science/paper/IX2K6MDJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.16405&json=true","fetch_graph":"https://pith.science/api/pith-number/IX2K6MDJWBAL333S5UPOJE4TN5/graph.json","fetch_events":"https://pith.science/api/pith-number/IX2K6MDJWBAL333S5UPOJE4TN5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IX2K6MDJWBAL333S5UPOJE4TN5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IX2K6MDJWBAL333S5UPOJE4TN5/action/storage_attestation","attest_author":"https://pith.science/pith/IX2K6MDJWBAL333S5UPOJE4TN5/action/author_attestation","sign_citation":"https://pith.science/pith/IX2K6MDJWBAL333S5UPOJE4TN5/action/citation_signature","submit_replication":"https://pith.science/pith/IX2K6MDJWBAL333S5UPOJE4TN5/action/replication_record"}},"created_at":"2026-07-05T07:40:43.710750+00:00","updated_at":"2026-07-05T07:40:43.710750+00:00"}