{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:H2IJ2QYI5J4EF2H2L2W4PZ5ETH","short_pith_number":"pith:H2IJ2QYI","schema_version":"1.0","canonical_sha256":"3e909d4308ea7842e8fa5eadc7e7a499e961d9aab87db032a7a89d12cb356fca","source":{"kind":"arxiv","id":"2311.15653","version":1},"attestation_state":"computed","paper":{"title":"MoDS: Model-oriented Data Selection for Instruction Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chengqing Zong, Jiajun Zhang, Qianlong Du","submitted_at":"2023-11-27T09:33:13Z","abstract_excerpt":"Instruction tuning has become the de facto method to equip large language models (LLMs) with the ability of following user instructions. Usually, hundreds of thousands or millions of instruction-following pairs are employed to fine-tune the foundation LLMs. Recently, some studies show that a small number of high-quality instruction data is enough. However, how to select appropriate instruction data for a given LLM is still an open problem. To address this problem, in this paper we present a model-oriented data selection (MoDS) approach, which selects instruction data based on a new criteria co"},"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":"2311.15653","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-27T09:33:13Z","cross_cats_sorted":[],"title_canon_sha256":"ba7746440afb8e39db77fae31c8f535cc584361789df19c3a47321e514ab3956","abstract_canon_sha256":"4e7359ae9a16c3b4ddf1fc09b0cea4065728e579e586c6e053e3b8138ff91330"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:17:07.580257Z","signature_b64":"d6Z/9EF/j+jb83rTEzSsASY1ETidXhTodChVbxnloZ3p5KrN3tS1Pf/JCq/cCbTogKZ0tfdeRFqdXBML8sieCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3e909d4308ea7842e8fa5eadc7e7a499e961d9aab87db032a7a89d12cb356fca","last_reissued_at":"2026-07-05T07:17:07.579517Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:17:07.579517Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MoDS: Model-oriented Data Selection for Instruction Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chengqing Zong, Jiajun Zhang, Qianlong Du","submitted_at":"2023-11-27T09:33:13Z","abstract_excerpt":"Instruction tuning has become the de facto method to equip large language models (LLMs) with the ability of following user instructions. Usually, hundreds of thousands or millions of instruction-following pairs are employed to fine-tune the foundation LLMs. Recently, some studies show that a small number of high-quality instruction data is enough. However, how to select appropriate instruction data for a given LLM is still an open problem. To address this problem, in this paper we present a model-oriented data selection (MoDS) approach, which selects instruction data based on a new criteria co"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.15653","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/2311.15653/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":"2311.15653","created_at":"2026-07-05T07:17:07.579595+00:00"},{"alias_kind":"arxiv_version","alias_value":"2311.15653v1","created_at":"2026-07-05T07:17:07.579595+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.15653","created_at":"2026-07-05T07:17:07.579595+00:00"},{"alias_kind":"pith_short_12","alias_value":"H2IJ2QYI5J4E","created_at":"2026-07-05T07:17:07.579595+00:00"},{"alias_kind":"pith_short_16","alias_value":"H2IJ2QYI5J4EF2H2","created_at":"2026-07-05T07:17:07.579595+00:00"},{"alias_kind":"pith_short_8","alias_value":"H2IJ2QYI","created_at":"2026-07-05T07:17:07.579595+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.10706","citing_title":"Unifying Data, Memory, and Compute Efficiency in LLM training: A Survey","ref_index":87,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30857","citing_title":"MADS: Model-Aware Diverse Core Set Selection for Instruction Tuning","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00400","citing_title":"Dynamic Proxy-Mixing: Transferring Replay Controllers from Small to Large Models for Continual Instruction Tuning","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2412.04468","citing_title":"NVILA: Efficient Frontier Visual Language Models","ref_index":85,"is_internal_anchor":false},{"citing_arxiv_id":"2502.10248","citing_title":"Step-Video-T2V Technical Report: The Practice, Challenges, and Future of Video Foundation Model","ref_index":105,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/H2IJ2QYI5J4EF2H2L2W4PZ5ETH","json":"https://pith.science/pith/H2IJ2QYI5J4EF2H2L2W4PZ5ETH.json","graph_json":"https://pith.science/api/pith-number/H2IJ2QYI5J4EF2H2L2W4PZ5ETH/graph.json","events_json":"https://pith.science/api/pith-number/H2IJ2QYI5J4EF2H2L2W4PZ5ETH/events.json","paper":"https://pith.science/paper/H2IJ2QYI"},"agent_actions":{"view_html":"https://pith.science/pith/H2IJ2QYI5J4EF2H2L2W4PZ5ETH","download_json":"https://pith.science/pith/H2IJ2QYI5J4EF2H2L2W4PZ5ETH.json","view_paper":"https://pith.science/paper/H2IJ2QYI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2311.15653&json=true","fetch_graph":"https://pith.science/api/pith-number/H2IJ2QYI5J4EF2H2L2W4PZ5ETH/graph.json","fetch_events":"https://pith.science/api/pith-number/H2IJ2QYI5J4EF2H2L2W4PZ5ETH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/H2IJ2QYI5J4EF2H2L2W4PZ5ETH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/H2IJ2QYI5J4EF2H2L2W4PZ5ETH/action/storage_attestation","attest_author":"https://pith.science/pith/H2IJ2QYI5J4EF2H2L2W4PZ5ETH/action/author_attestation","sign_citation":"https://pith.science/pith/H2IJ2QYI5J4EF2H2L2W4PZ5ETH/action/citation_signature","submit_replication":"https://pith.science/pith/H2IJ2QYI5J4EF2H2L2W4PZ5ETH/action/replication_record"}},"created_at":"2026-07-05T07:17:07.579595+00:00","updated_at":"2026-07-05T07:17:07.579595+00:00"}