{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:PYS3C5I6ZSGQJCQPLQ4I3RL6OO","short_pith_number":"pith:PYS3C5I6","schema_version":"1.0","canonical_sha256":"7e25b1751ecc8d048a0f5c388dc57e738bdd1ab33967a79d05eb4e941401c886","source":{"kind":"arxiv","id":"2407.12259","version":2},"attestation_state":"computed","paper":{"title":"On the Feasibility of In-Context Probing for Data Attribution","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Aditi Raghunathan, Cathy Jiao, Chenyan Xiong, Gary Gao","submitted_at":"2024-07-17T02:06:56Z","abstract_excerpt":"Data attribution methods are used to measure the contribution of training data towards model outputs, and have several important applications in areas such as dataset curation and model interpretability. However, many standard data attribution methods, such as influence functions, utilize model gradients and are computationally expensive. In our paper, we show in-context probing (ICP) -- prompting a LLM -- can serve as a fast proxy for gradient-based data attribution for data selection under conditions contingent on data similarity. We study this connection empirically on standard NLP tasks, a"},"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":"2407.12259","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-17T02:06:56Z","cross_cats_sorted":[],"title_canon_sha256":"fcc907895c0708b253091bda678db89524a4657a08ac05f81ee41b7813bcd891","abstract_canon_sha256":"5bd331d4e26792c6c24fc5cc650f37e345ca1c6038edb8a8096fef018cc2ef27"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:12:18.092815Z","signature_b64":"LzGPf3FVGjUgbx1VsN7lUB+8aBs2hOo5s0QyYylFWpcbkg+KKISXVLBh1EqNJ5O0ZvkTkxQOxpB+QOscSw2ODg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7e25b1751ecc8d048a0f5c388dc57e738bdd1ab33967a79d05eb4e941401c886","last_reissued_at":"2026-07-05T10:12:18.092307Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:12:18.092307Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On the Feasibility of In-Context Probing for Data Attribution","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Aditi Raghunathan, Cathy Jiao, Chenyan Xiong, Gary Gao","submitted_at":"2024-07-17T02:06:56Z","abstract_excerpt":"Data attribution methods are used to measure the contribution of training data towards model outputs, and have several important applications in areas such as dataset curation and model interpretability. However, many standard data attribution methods, such as influence functions, utilize model gradients and are computationally expensive. In our paper, we show in-context probing (ICP) -- prompting a LLM -- can serve as a fast proxy for gradient-based data attribution for data selection under conditions contingent on data similarity. We study this connection empirically on standard NLP tasks, a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.12259","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/2407.12259/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":"2407.12259","created_at":"2026-07-05T10:12:18.092379+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.12259v2","created_at":"2026-07-05T10:12:18.092379+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.12259","created_at":"2026-07-05T10:12:18.092379+00:00"},{"alias_kind":"pith_short_12","alias_value":"PYS3C5I6ZSGQ","created_at":"2026-07-05T10:12:18.092379+00:00"},{"alias_kind":"pith_short_16","alias_value":"PYS3C5I6ZSGQJCQP","created_at":"2026-07-05T10:12:18.092379+00:00"},{"alias_kind":"pith_short_8","alias_value":"PYS3C5I6","created_at":"2026-07-05T10:12:18.092379+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/PYS3C5I6ZSGQJCQPLQ4I3RL6OO","json":"https://pith.science/pith/PYS3C5I6ZSGQJCQPLQ4I3RL6OO.json","graph_json":"https://pith.science/api/pith-number/PYS3C5I6ZSGQJCQPLQ4I3RL6OO/graph.json","events_json":"https://pith.science/api/pith-number/PYS3C5I6ZSGQJCQPLQ4I3RL6OO/events.json","paper":"https://pith.science/paper/PYS3C5I6"},"agent_actions":{"view_html":"https://pith.science/pith/PYS3C5I6ZSGQJCQPLQ4I3RL6OO","download_json":"https://pith.science/pith/PYS3C5I6ZSGQJCQPLQ4I3RL6OO.json","view_paper":"https://pith.science/paper/PYS3C5I6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.12259&json=true","fetch_graph":"https://pith.science/api/pith-number/PYS3C5I6ZSGQJCQPLQ4I3RL6OO/graph.json","fetch_events":"https://pith.science/api/pith-number/PYS3C5I6ZSGQJCQPLQ4I3RL6OO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PYS3C5I6ZSGQJCQPLQ4I3RL6OO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PYS3C5I6ZSGQJCQPLQ4I3RL6OO/action/storage_attestation","attest_author":"https://pith.science/pith/PYS3C5I6ZSGQJCQPLQ4I3RL6OO/action/author_attestation","sign_citation":"https://pith.science/pith/PYS3C5I6ZSGQJCQPLQ4I3RL6OO/action/citation_signature","submit_replication":"https://pith.science/pith/PYS3C5I6ZSGQJCQPLQ4I3RL6OO/action/replication_record"}},"created_at":"2026-07-05T10:12:18.092379+00:00","updated_at":"2026-07-05T10:12:18.092379+00:00"}