{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:5IHBJRLSXWPDHLSDI4R6OTZWGS","short_pith_number":"pith:5IHBJRLS","schema_version":"1.0","canonical_sha256":"ea0e14c572bd9e33ae434723e74f3634ab0f130d9ea316fa69a4c3b67823674e","source":{"kind":"arxiv","id":"2412.06293","version":2},"attestation_state":"computed","paper":{"title":"Mastering Collaborative Multi-modal Data Selection: A Focus on Informativeness, Uniqueness, and Representativeness","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bosheng Qin, Juncheng Li, Qifan Yu, Siliang Tang, Wenqiao Zhang, Yang Wu, Yueting Zhuang, Yunfei Li, Zhebei Shen, Zhongqi Yue","submitted_at":"2024-12-09T08:36:10Z","abstract_excerpt":"Instruction tuning fine-tunes pre-trained Multi-modal Large Language Models (MLLMs) to handle real-world tasks. However, the rapid expansion of visual instruction datasets introduces data redundancy, leading to excessive computational costs. We propose a collaborative framework, DataTailor, which leverages three key principles--informativeness, uniqueness, and representativeness--for effective data selection. We argue that a valuable sample should be informative of the task, non-redundant, and represent the sample distribution (i.e., not an outlier). We further propose practical ways to score "},"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":"2412.06293","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-09T08:36:10Z","cross_cats_sorted":[],"title_canon_sha256":"0ce614553932c20c2d9157b09ae901f284a57321c486b7aa868e9f2a10a9557c","abstract_canon_sha256":"0f8aeb15856af288836da3c5a7e946bbb6a3817c92878f90c45fe720149ffb3c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:53:43.026019Z","signature_b64":"jxSiBzvz0GTkukvmbFkn2lLwT7F7yMMYJS1nkfnYR9HbgdYmK2fpHewFyi3edrpX2Y/rlh/swNvkxFoC8sSuCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ea0e14c572bd9e33ae434723e74f3634ab0f130d9ea316fa69a4c3b67823674e","last_reissued_at":"2026-07-05T11:53:43.025524Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:53:43.025524Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mastering Collaborative Multi-modal Data Selection: A Focus on Informativeness, Uniqueness, and Representativeness","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bosheng Qin, Juncheng Li, Qifan Yu, Siliang Tang, Wenqiao Zhang, Yang Wu, Yueting Zhuang, Yunfei Li, Zhebei Shen, Zhongqi Yue","submitted_at":"2024-12-09T08:36:10Z","abstract_excerpt":"Instruction tuning fine-tunes pre-trained Multi-modal Large Language Models (MLLMs) to handle real-world tasks. However, the rapid expansion of visual instruction datasets introduces data redundancy, leading to excessive computational costs. We propose a collaborative framework, DataTailor, which leverages three key principles--informativeness, uniqueness, and representativeness--for effective data selection. We argue that a valuable sample should be informative of the task, non-redundant, and represent the sample distribution (i.e., not an outlier). We further propose practical ways to score "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.06293","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/2412.06293/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":"2412.06293","created_at":"2026-07-05T11:53:43.025577+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.06293v2","created_at":"2026-07-05T11:53:43.025577+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.06293","created_at":"2026-07-05T11:53:43.025577+00:00"},{"alias_kind":"pith_short_12","alias_value":"5IHBJRLSXWPD","created_at":"2026-07-05T11:53:43.025577+00:00"},{"alias_kind":"pith_short_16","alias_value":"5IHBJRLSXWPDHLSD","created_at":"2026-07-05T11:53:43.025577+00:00"},{"alias_kind":"pith_short_8","alias_value":"5IHBJRLS","created_at":"2026-07-05T11:53:43.025577+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.00068","citing_title":"SafeBuild-Bench: A Temporal-Robust Construction Safety Benchmark with Graph-Enhanced Data Mining","ref_index":45,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5IHBJRLSXWPDHLSDI4R6OTZWGS","json":"https://pith.science/pith/5IHBJRLSXWPDHLSDI4R6OTZWGS.json","graph_json":"https://pith.science/api/pith-number/5IHBJRLSXWPDHLSDI4R6OTZWGS/graph.json","events_json":"https://pith.science/api/pith-number/5IHBJRLSXWPDHLSDI4R6OTZWGS/events.json","paper":"https://pith.science/paper/5IHBJRLS"},"agent_actions":{"view_html":"https://pith.science/pith/5IHBJRLSXWPDHLSDI4R6OTZWGS","download_json":"https://pith.science/pith/5IHBJRLSXWPDHLSDI4R6OTZWGS.json","view_paper":"https://pith.science/paper/5IHBJRLS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.06293&json=true","fetch_graph":"https://pith.science/api/pith-number/5IHBJRLSXWPDHLSDI4R6OTZWGS/graph.json","fetch_events":"https://pith.science/api/pith-number/5IHBJRLSXWPDHLSDI4R6OTZWGS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5IHBJRLSXWPDHLSDI4R6OTZWGS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5IHBJRLSXWPDHLSDI4R6OTZWGS/action/storage_attestation","attest_author":"https://pith.science/pith/5IHBJRLSXWPDHLSDI4R6OTZWGS/action/author_attestation","sign_citation":"https://pith.science/pith/5IHBJRLSXWPDHLSDI4R6OTZWGS/action/citation_signature","submit_replication":"https://pith.science/pith/5IHBJRLSXWPDHLSDI4R6OTZWGS/action/replication_record"}},"created_at":"2026-07-05T11:53:43.025577+00:00","updated_at":"2026-07-05T11:53:43.025577+00:00"}