{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XTUJ3Z44BXB4GA2YE3EY67ZWV2","short_pith_number":"pith:XTUJ3Z44","schema_version":"1.0","canonical_sha256":"bce89de79c0dc3c3035826c98f7f36aebc33766352027af8a50c88f2c5a60122","source":{"kind":"arxiv","id":"2510.14261","version":2},"attestation_state":"computed","paper":{"title":"Rewriting History: A Recipe for Interventional Analyses to Study Data Effects on Model Behavior","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hila Gonen, Noah A. Smith, Rahul Nadkarni, Yanai Elazar","submitted_at":"2025-10-16T03:22:48Z","abstract_excerpt":"We present an experimental recipe for studying the relationship between training data and language model (LM) behavior. We outline steps for intervening on data batches -- i.e., ``rewriting history'' -- and then retraining model checkpoints over that data to test hypotheses relating data to behavior. Our recipe breaks down such an intervention into stages that include selecting evaluation items from a benchmark that measures model behavior, matching relevant documents to those items, and modifying those documents before retraining and measuring the effects. We demonstrate the utility of our re"},"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":"2510.14261","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-10-16T03:22:48Z","cross_cats_sorted":[],"title_canon_sha256":"b4ff8602b5b2fbbacbad94fae8875f6b4bd85727124387169f61caa1eb94831b","abstract_canon_sha256":"a598fcf17f1c2e17332245f5616a2807fc841cd104ee656504e40ad5d662c364"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:05:00.242205Z","signature_b64":"Ve6kE3a/aCAjyYflCY1rqJwMEpBzR/kgq6hE98g0mUzGH3No5MXrbwxBq59Eh4yHjo6ZkiQpueMvQ+nnzw03Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bce89de79c0dc3c3035826c98f7f36aebc33766352027af8a50c88f2c5a60122","last_reissued_at":"2026-05-20T01:05:00.241329Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:05:00.241329Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rewriting History: A Recipe for Interventional Analyses to Study Data Effects on Model Behavior","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hila Gonen, Noah A. Smith, Rahul Nadkarni, Yanai Elazar","submitted_at":"2025-10-16T03:22:48Z","abstract_excerpt":"We present an experimental recipe for studying the relationship between training data and language model (LM) behavior. We outline steps for intervening on data batches -- i.e., ``rewriting history'' -- and then retraining model checkpoints over that data to test hypotheses relating data to behavior. Our recipe breaks down such an intervention into stages that include selecting evaluation items from a benchmark that measures model behavior, matching relevant documents to those items, and modifying those documents before retraining and measuring the effects. We demonstrate the utility of our re"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.14261","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/2510.14261/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":"2510.14261","created_at":"2026-05-20T01:05:00.241446+00:00"},{"alias_kind":"arxiv_version","alias_value":"2510.14261v2","created_at":"2026-05-20T01:05:00.241446+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.14261","created_at":"2026-05-20T01:05:00.241446+00:00"},{"alias_kind":"pith_short_12","alias_value":"XTUJ3Z44BXB4","created_at":"2026-05-20T01:05:00.241446+00:00"},{"alias_kind":"pith_short_16","alias_value":"XTUJ3Z44BXB4GA2Y","created_at":"2026-05-20T01:05:00.241446+00:00"},{"alias_kind":"pith_short_8","alias_value":"XTUJ3Z44","created_at":"2026-05-20T01:05:00.241446+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/XTUJ3Z44BXB4GA2YE3EY67ZWV2","json":"https://pith.science/pith/XTUJ3Z44BXB4GA2YE3EY67ZWV2.json","graph_json":"https://pith.science/api/pith-number/XTUJ3Z44BXB4GA2YE3EY67ZWV2/graph.json","events_json":"https://pith.science/api/pith-number/XTUJ3Z44BXB4GA2YE3EY67ZWV2/events.json","paper":"https://pith.science/paper/XTUJ3Z44"},"agent_actions":{"view_html":"https://pith.science/pith/XTUJ3Z44BXB4GA2YE3EY67ZWV2","download_json":"https://pith.science/pith/XTUJ3Z44BXB4GA2YE3EY67ZWV2.json","view_paper":"https://pith.science/paper/XTUJ3Z44","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2510.14261&json=true","fetch_graph":"https://pith.science/api/pith-number/XTUJ3Z44BXB4GA2YE3EY67ZWV2/graph.json","fetch_events":"https://pith.science/api/pith-number/XTUJ3Z44BXB4GA2YE3EY67ZWV2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XTUJ3Z44BXB4GA2YE3EY67ZWV2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XTUJ3Z44BXB4GA2YE3EY67ZWV2/action/storage_attestation","attest_author":"https://pith.science/pith/XTUJ3Z44BXB4GA2YE3EY67ZWV2/action/author_attestation","sign_citation":"https://pith.science/pith/XTUJ3Z44BXB4GA2YE3EY67ZWV2/action/citation_signature","submit_replication":"https://pith.science/pith/XTUJ3Z44BXB4GA2YE3EY67ZWV2/action/replication_record"}},"created_at":"2026-05-20T01:05:00.241446+00:00","updated_at":"2026-05-20T01:05:00.241446+00:00"}