{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:YPQ4SC7LW5UQHDRXPNAYOSGI4R","short_pith_number":"pith:YPQ4SC7L","schema_version":"1.0","canonical_sha256":"c3e1c90bebb769038e377b418748c8e45eb74e5ed4ffba254d5aa5efdb7bf5be","source":{"kind":"arxiv","id":"2608.13515","version":1},"attestation_state":"computed","paper":{"title":"Measuring Task-Agnostic Training Data Influence Across Language Model Pretraining","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chaoran Liu, Daisuke Kawahara, Hirokazu Kiyomaru, Masaru Isonuma, Max M\\\"uller-Eberstein, Takashi Kodama, Yusuke Miyao, Yusuke Oda, Yuto Nishida","submitted_at":"2026-08-13T17:36:49Z","abstract_excerpt":"Measuring training data influence consistently across language model pretraining is challenging. It is difficult to select downstream tasks or validation sets representative of a model's general capabilities, and reliance on task performance at intermediate checkpoints complicates comparisons across training. We propose a measure of training data influence that does not require selecting a downstream task or validation set as the attribution target. Specifically, we define an example's influence by how much its gradient update reduces the squared distance to the final parameters of a given pre"},"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":"2608.13515","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-13T17:36:49Z","cross_cats_sorted":[],"title_canon_sha256":"d905870b307728bfff727db706ea386608ef4fac3e8e508c7526d48307a9afe7","abstract_canon_sha256":"770420108413d1db8c46364c1cadcaab90e086d658da37967fad82fd865ddca0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-14T01:24:55.853349Z","signature_b64":"EnCQYeZnA5ckkmFBqsgCd5OBwR8l4cAqAT4eODnGzPRh7wxjjybafwilCFh6LRKnWIxSDkvIeTDemJklUzPfBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c3e1c90bebb769038e377b418748c8e45eb74e5ed4ffba254d5aa5efdb7bf5be","last_reissued_at":"2026-08-14T01:24:55.851101Z","signature_status":"signed_v1","first_computed_at":"2026-08-14T01:24:55.851101Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Measuring Task-Agnostic Training Data Influence Across Language Model Pretraining","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chaoran Liu, Daisuke Kawahara, Hirokazu Kiyomaru, Masaru Isonuma, Max M\\\"uller-Eberstein, Takashi Kodama, Yusuke Miyao, Yusuke Oda, Yuto Nishida","submitted_at":"2026-08-13T17:36:49Z","abstract_excerpt":"Measuring training data influence consistently across language model pretraining is challenging. It is difficult to select downstream tasks or validation sets representative of a model's general capabilities, and reliance on task performance at intermediate checkpoints complicates comparisons across training. We propose a measure of training data influence that does not require selecting a downstream task or validation set as the attribution target. Specifically, we define an example's influence by how much its gradient update reduces the squared distance to the final parameters of a given pre"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.13515","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/2608.13515/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":"2608.13515","created_at":"2026-08-14T01:24:55.852631+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.13515v1","created_at":"2026-08-14T01:24:55.852631+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.13515","created_at":"2026-08-14T01:24:55.852631+00:00"},{"alias_kind":"pith_short_12","alias_value":"YPQ4SC7LW5UQ","created_at":"2026-08-14T01:24:55.852631+00:00"},{"alias_kind":"pith_short_16","alias_value":"YPQ4SC7LW5UQHDRX","created_at":"2026-08-14T01:24:55.852631+00:00"},{"alias_kind":"pith_short_8","alias_value":"YPQ4SC7L","created_at":"2026-08-14T01:24:55.852631+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/YPQ4SC7LW5UQHDRXPNAYOSGI4R","json":"https://pith.science/pith/YPQ4SC7LW5UQHDRXPNAYOSGI4R.json","graph_json":"https://pith.science/api/pith-number/YPQ4SC7LW5UQHDRXPNAYOSGI4R/graph.json","events_json":"https://pith.science/api/pith-number/YPQ4SC7LW5UQHDRXPNAYOSGI4R/events.json","paper":"https://pith.science/paper/YPQ4SC7L"},"agent_actions":{"view_html":"https://pith.science/pith/YPQ4SC7LW5UQHDRXPNAYOSGI4R","download_json":"https://pith.science/pith/YPQ4SC7LW5UQHDRXPNAYOSGI4R.json","view_paper":"https://pith.science/paper/YPQ4SC7L","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.13515&json=true","fetch_graph":"https://pith.science/api/pith-number/YPQ4SC7LW5UQHDRXPNAYOSGI4R/graph.json","fetch_events":"https://pith.science/api/pith-number/YPQ4SC7LW5UQHDRXPNAYOSGI4R/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YPQ4SC7LW5UQHDRXPNAYOSGI4R/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YPQ4SC7LW5UQHDRXPNAYOSGI4R/action/storage_attestation","attest_author":"https://pith.science/pith/YPQ4SC7LW5UQHDRXPNAYOSGI4R/action/author_attestation","sign_citation":"https://pith.science/pith/YPQ4SC7LW5UQHDRXPNAYOSGI4R/action/citation_signature","submit_replication":"https://pith.science/pith/YPQ4SC7LW5UQHDRXPNAYOSGI4R/action/replication_record"}},"created_at":"2026-08-14T01:24:55.852631+00:00","updated_at":"2026-08-14T01:24:55.852631+00:00"}