{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6V6WFWW4ROSUZIRLTDUKEUAVKD","short_pith_number":"pith:6V6WFWW4","schema_version":"1.0","canonical_sha256":"f57d62dadc8ba54ca22b98e8a2501550cbce8e1651f9220ce5e11554b97e75a6","source":{"kind":"arxiv","id":"2401.13586","version":4},"attestation_state":"computed","paper":{"title":"Instruction Fine-Tuning: Does Prompt Loss Matter?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Mathew Huerta-Enochian, Seung Yong Ko","submitted_at":"2024-01-24T16:51:23Z","abstract_excerpt":"We present a novel study analyzing the effects of various prompt loss token weights (PLW) for supervised instruction fine-tuning (SIFT). While prompt-masking (PLW = 0) is common for SIFT, some fine-tuning APIs support fractional PLWs and suggest that using a small non-zero PLW can help stabilize learning when fine-tuning on short-completion data. However, there has never been a study confirming this claim, and OpenAI, a major cloud-based SIFT provider, recently removed this parameter from their fine-tuning API. We found that performance of models fine-tuned on short-completion data had a stati"},"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":"2401.13586","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-01-24T16:51:23Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"5c4be94765f2bc34c25423262c2895de68dbc6b187b37e909d3fdc1ce8b6480d","abstract_canon_sha256":"7f597be25c1b9ed51bddafbc42668a66748f588b5ccb1b2cb1cf42a39a86341e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:19:53.717122Z","signature_b64":"lXmCDPlxUjoiS/4Qr/x+I9wT1OZZUnZWupII2i9zeuPjtf22VZhkG1AaanstLEzfs0E2fDtg9hckc6Uf5kADDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f57d62dadc8ba54ca22b98e8a2501550cbce8e1651f9220ce5e11554b97e75a6","last_reissued_at":"2026-07-05T09:19:53.716620Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:19:53.716620Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Instruction Fine-Tuning: Does Prompt Loss Matter?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Mathew Huerta-Enochian, Seung Yong Ko","submitted_at":"2024-01-24T16:51:23Z","abstract_excerpt":"We present a novel study analyzing the effects of various prompt loss token weights (PLW) for supervised instruction fine-tuning (SIFT). While prompt-masking (PLW = 0) is common for SIFT, some fine-tuning APIs support fractional PLWs and suggest that using a small non-zero PLW can help stabilize learning when fine-tuning on short-completion data. However, there has never been a study confirming this claim, and OpenAI, a major cloud-based SIFT provider, recently removed this parameter from their fine-tuning API. We found that performance of models fine-tuned on short-completion data had a stati"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.13586","kind":"arxiv","version":4},"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/2401.13586/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":"2401.13586","created_at":"2026-07-05T09:19:53.716679+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.13586v4","created_at":"2026-07-05T09:19:53.716679+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.13586","created_at":"2026-07-05T09:19:53.716679+00:00"},{"alias_kind":"pith_short_12","alias_value":"6V6WFWW4ROSU","created_at":"2026-07-05T09:19:53.716679+00:00"},{"alias_kind":"pith_short_16","alias_value":"6V6WFWW4ROSUZIRL","created_at":"2026-07-05T09:19:53.716679+00:00"},{"alias_kind":"pith_short_8","alias_value":"6V6WFWW4","created_at":"2026-07-05T09:19:53.716679+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.19528","citing_title":"Techniques for Peak Memory Reduction for LoRA Fine-tuning of LLMs on Edge Devices","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2601.23206","citing_title":"High-quality generation of dynamic game content via small language models: A proof of concept","ref_index":11,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6V6WFWW4ROSUZIRLTDUKEUAVKD","json":"https://pith.science/pith/6V6WFWW4ROSUZIRLTDUKEUAVKD.json","graph_json":"https://pith.science/api/pith-number/6V6WFWW4ROSUZIRLTDUKEUAVKD/graph.json","events_json":"https://pith.science/api/pith-number/6V6WFWW4ROSUZIRLTDUKEUAVKD/events.json","paper":"https://pith.science/paper/6V6WFWW4"},"agent_actions":{"view_html":"https://pith.science/pith/6V6WFWW4ROSUZIRLTDUKEUAVKD","download_json":"https://pith.science/pith/6V6WFWW4ROSUZIRLTDUKEUAVKD.json","view_paper":"https://pith.science/paper/6V6WFWW4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.13586&json=true","fetch_graph":"https://pith.science/api/pith-number/6V6WFWW4ROSUZIRLTDUKEUAVKD/graph.json","fetch_events":"https://pith.science/api/pith-number/6V6WFWW4ROSUZIRLTDUKEUAVKD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6V6WFWW4ROSUZIRLTDUKEUAVKD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6V6WFWW4ROSUZIRLTDUKEUAVKD/action/storage_attestation","attest_author":"https://pith.science/pith/6V6WFWW4ROSUZIRLTDUKEUAVKD/action/author_attestation","sign_citation":"https://pith.science/pith/6V6WFWW4ROSUZIRLTDUKEUAVKD/action/citation_signature","submit_replication":"https://pith.science/pith/6V6WFWW4ROSUZIRLTDUKEUAVKD/action/replication_record"}},"created_at":"2026-07-05T09:19:53.716679+00:00","updated_at":"2026-07-05T09:19:53.716679+00:00"}