{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:I5FD5HZ2NBTGIS7XCJKTBZ6P5F","short_pith_number":"pith:I5FD5HZ2","schema_version":"1.0","canonical_sha256":"474a3e9f3a6866644bf7125530e7cfe95ae91e4dad4938717d2e4badde599fe7","source":{"kind":"arxiv","id":"2505.05327","version":2},"attestation_state":"computed","paper":{"title":"RICo: Refined In-Context Contribution for Automatic Instruction-Tuning Data Selection","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Fangwei Zhu, Linli Yao, Qingxiu Dong, Yixin Yang, Zhifang Sui","submitted_at":"2025-05-08T15:17:37Z","abstract_excerpt":"Data selection for instruction tuning is crucial for improving the performance of large language models (LLMs) while reducing training costs. In this paper, we propose Refined Contribution Measurement with In-Context Learning (RICo), a novel gradient-free method that quantifies the fine-grained contribution of individual samples to both task-level and global-level model performance. RICo enables more accurate identification of high-contribution data, leading to better instruction tuning. We further introduce a lightweight selection paradigm trained on RICo scores, enabling scalable data select"},"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":"2505.05327","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-08T15:17:37Z","cross_cats_sorted":[],"title_canon_sha256":"c57ec4872e3f949ca7372a327aca28d8a124e574e8c291538961654983a68ea5","abstract_canon_sha256":"e3d6e0addde383bc89e5ef08a61c41ef9d8f7706e3452cb423c1274f6de4e1f0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:04:50.727588Z","signature_b64":"xu3nsJOSq9M5ET2b+Z0piU84Be8GAUNnK+vv4uoszYk+fiymrMIQHPl0+CkzOyG/h/7Pys+K1HaAXNUacv1BBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"474a3e9f3a6866644bf7125530e7cfe95ae91e4dad4938717d2e4badde599fe7","last_reissued_at":"2026-07-05T11:04:50.727019Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:04:50.727019Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"RICo: Refined In-Context Contribution for Automatic Instruction-Tuning Data Selection","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Fangwei Zhu, Linli Yao, Qingxiu Dong, Yixin Yang, Zhifang Sui","submitted_at":"2025-05-08T15:17:37Z","abstract_excerpt":"Data selection for instruction tuning is crucial for improving the performance of large language models (LLMs) while reducing training costs. In this paper, we propose Refined Contribution Measurement with In-Context Learning (RICo), a novel gradient-free method that quantifies the fine-grained contribution of individual samples to both task-level and global-level model performance. RICo enables more accurate identification of high-contribution data, leading to better instruction tuning. We further introduce a lightweight selection paradigm trained on RICo scores, enabling scalable data select"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.05327","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/2505.05327/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":"2505.05327","created_at":"2026-07-05T11:04:50.727087+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.05327v2","created_at":"2026-07-05T11:04:50.727087+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.05327","created_at":"2026-07-05T11:04:50.727087+00:00"},{"alias_kind":"pith_short_12","alias_value":"I5FD5HZ2NBTG","created_at":"2026-07-05T11:04:50.727087+00:00"},{"alias_kind":"pith_short_16","alias_value":"I5FD5HZ2NBTGIS7X","created_at":"2026-07-05T11:04:50.727087+00:00"},{"alias_kind":"pith_short_8","alias_value":"I5FD5HZ2","created_at":"2026-07-05T11:04:50.727087+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/I5FD5HZ2NBTGIS7XCJKTBZ6P5F","json":"https://pith.science/pith/I5FD5HZ2NBTGIS7XCJKTBZ6P5F.json","graph_json":"https://pith.science/api/pith-number/I5FD5HZ2NBTGIS7XCJKTBZ6P5F/graph.json","events_json":"https://pith.science/api/pith-number/I5FD5HZ2NBTGIS7XCJKTBZ6P5F/events.json","paper":"https://pith.science/paper/I5FD5HZ2"},"agent_actions":{"view_html":"https://pith.science/pith/I5FD5HZ2NBTGIS7XCJKTBZ6P5F","download_json":"https://pith.science/pith/I5FD5HZ2NBTGIS7XCJKTBZ6P5F.json","view_paper":"https://pith.science/paper/I5FD5HZ2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.05327&json=true","fetch_graph":"https://pith.science/api/pith-number/I5FD5HZ2NBTGIS7XCJKTBZ6P5F/graph.json","fetch_events":"https://pith.science/api/pith-number/I5FD5HZ2NBTGIS7XCJKTBZ6P5F/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/I5FD5HZ2NBTGIS7XCJKTBZ6P5F/action/timestamp_anchor","attest_storage":"https://pith.science/pith/I5FD5HZ2NBTGIS7XCJKTBZ6P5F/action/storage_attestation","attest_author":"https://pith.science/pith/I5FD5HZ2NBTGIS7XCJKTBZ6P5F/action/author_attestation","sign_citation":"https://pith.science/pith/I5FD5HZ2NBTGIS7XCJKTBZ6P5F/action/citation_signature","submit_replication":"https://pith.science/pith/I5FD5HZ2NBTGIS7XCJKTBZ6P5F/action/replication_record"}},"created_at":"2026-07-05T11:04:50.727087+00:00","updated_at":"2026-07-05T11:04:50.727087+00:00"}