{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:MSNILLIYUB6YZZZ3GPQEZR4FTU","short_pith_number":"pith:MSNILLIY","schema_version":"1.0","canonical_sha256":"649a85ad18a07d8ce73b33e04cc7859d32ab8d6fbf0cef732f8c7f1042a0875f","source":{"kind":"arxiv","id":"2402.12501","version":1},"attestation_state":"computed","paper":{"title":"Your Vision-Language Model Itself Is a Strong Filter: Towards High-Quality Instruction Tuning with Data Selection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chenxi Liu, Guodong Liu, Heng Huang, Junfeng Guo, Lichang Chen, Qi He, Ruibo Chen, Tianyi Xiong, Yihan Wu","submitted_at":"2024-02-19T20:08:48Z","abstract_excerpt":"Data selection in instruction tuning emerges as a pivotal process for acquiring high-quality data and training instruction-following large language models (LLMs), but it is still a new and unexplored research area for vision-language models (VLMs). Existing data selection approaches on LLMs either rely on single unreliable scores, or use downstream tasks for selection, which is time-consuming and can lead to potential over-fitting on the chosen evaluation datasets. To address this challenge, we introduce a novel dataset selection method, Self-Filter, that utilizes the VLM itself as a filter. T"},"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":"2402.12501","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-19T20:08:48Z","cross_cats_sorted":[],"title_canon_sha256":"c0cfd2039903060e43d9236a3aa6e537f23bf9ff451b78c71f0b6d2fa495096e","abstract_canon_sha256":"c508331177c0fcfdeb3819c71f31712d8e96056530820a779596b599df9096ac"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:47:12.215935Z","signature_b64":"Pw9ujIIhVMTt0SbgxPLwFch9b8DXjhnvi2pogwrmRtxBtLMXC7oqkdaPdQLv/B0C9yncGUZW5LNX1doDi8HiCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"649a85ad18a07d8ce73b33e04cc7859d32ab8d6fbf0cef732f8c7f1042a0875f","last_reissued_at":"2026-07-05T07:47:12.215549Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:47:12.215549Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Your Vision-Language Model Itself Is a Strong Filter: Towards High-Quality Instruction Tuning with Data Selection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chenxi Liu, Guodong Liu, Heng Huang, Junfeng Guo, Lichang Chen, Qi He, Ruibo Chen, Tianyi Xiong, Yihan Wu","submitted_at":"2024-02-19T20:08:48Z","abstract_excerpt":"Data selection in instruction tuning emerges as a pivotal process for acquiring high-quality data and training instruction-following large language models (LLMs), but it is still a new and unexplored research area for vision-language models (VLMs). Existing data selection approaches on LLMs either rely on single unreliable scores, or use downstream tasks for selection, which is time-consuming and can lead to potential over-fitting on the chosen evaluation datasets. To address this challenge, we introduce a novel dataset selection method, Self-Filter, that utilizes the VLM itself as a filter. T"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.12501","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/2402.12501/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":"2402.12501","created_at":"2026-07-05T07:47:12.215604+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.12501v1","created_at":"2026-07-05T07:47:12.215604+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.12501","created_at":"2026-07-05T07:47:12.215604+00:00"},{"alias_kind":"pith_short_12","alias_value":"MSNILLIYUB6Y","created_at":"2026-07-05T07:47:12.215604+00:00"},{"alias_kind":"pith_short_16","alias_value":"MSNILLIYUB6YZZZ3","created_at":"2026-07-05T07:47:12.215604+00:00"},{"alias_kind":"pith_short_8","alias_value":"MSNILLIY","created_at":"2026-07-05T07:47:12.215604+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.03879","citing_title":"CL3DOR: Contrastive Learning for 3D Large Multimodal Models via Odds Ratio on High-Resolution Point Clouds","ref_index":11,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MSNILLIYUB6YZZZ3GPQEZR4FTU","json":"https://pith.science/pith/MSNILLIYUB6YZZZ3GPQEZR4FTU.json","graph_json":"https://pith.science/api/pith-number/MSNILLIYUB6YZZZ3GPQEZR4FTU/graph.json","events_json":"https://pith.science/api/pith-number/MSNILLIYUB6YZZZ3GPQEZR4FTU/events.json","paper":"https://pith.science/paper/MSNILLIY"},"agent_actions":{"view_html":"https://pith.science/pith/MSNILLIYUB6YZZZ3GPQEZR4FTU","download_json":"https://pith.science/pith/MSNILLIYUB6YZZZ3GPQEZR4FTU.json","view_paper":"https://pith.science/paper/MSNILLIY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.12501&json=true","fetch_graph":"https://pith.science/api/pith-number/MSNILLIYUB6YZZZ3GPQEZR4FTU/graph.json","fetch_events":"https://pith.science/api/pith-number/MSNILLIYUB6YZZZ3GPQEZR4FTU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MSNILLIYUB6YZZZ3GPQEZR4FTU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MSNILLIYUB6YZZZ3GPQEZR4FTU/action/storage_attestation","attest_author":"https://pith.science/pith/MSNILLIYUB6YZZZ3GPQEZR4FTU/action/author_attestation","sign_citation":"https://pith.science/pith/MSNILLIYUB6YZZZ3GPQEZR4FTU/action/citation_signature","submit_replication":"https://pith.science/pith/MSNILLIYUB6YZZZ3GPQEZR4FTU/action/replication_record"}},"created_at":"2026-07-05T07:47:12.215604+00:00","updated_at":"2026-07-05T07:47:12.215604+00:00"}