{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:HII5HMGUTHR6NY7DP2C3HFGD2G","short_pith_number":"pith:HII5HMGU","schema_version":"1.0","canonical_sha256":"3a11d3b0d499e3e6e3e37e85b394c3d1ba751a41dac3b22edf518e54294c32c1","source":{"kind":"arxiv","id":"2012.15781","version":2},"attestation_state":"computed","paper":{"title":"FastIF: Scalable Influence Functions for Efficient Model Interpretation and Debugging","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Caiming Xiong, Han Guo, Mohit Bansal, Nazneen Fatema Rajani, Peter Hase","submitted_at":"2020-12-31T18:02:34Z","abstract_excerpt":"Influence functions approximate the \"influences\" of training data-points for test predictions and have a wide variety of applications. Despite the popularity, their computational cost does not scale well with model and training data size. We present FastIF, a set of simple modifications to influence functions that significantly improves their run-time. We use k-Nearest Neighbors (kNN) to narrow the search space down to a subset of good candidate data points, identify the configurations that best balance the speed-quality trade-off in estimating the inverse Hessian-vector product, and introduce"},"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":"2012.15781","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-12-31T18:02:34Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"f89279a882c266b45c0c7c73979e739f1a8b6d223d43fa1ec586fd9c6e1ad4ab","abstract_canon_sha256":"332e6971072b4202090af0183e814965e1925aeff9bdfb65c68529407629fdd5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:13:01.697437Z","signature_b64":"swQBU+UxFWQQ/oX53VDApc0cSObjhanPAzJ57B2w7iuYX7qvUN4ar9YXtuIhms8crdCXc1W/5KTApMUf2W22DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3a11d3b0d499e3e6e3e37e85b394c3d1ba751a41dac3b22edf518e54294c32c1","last_reissued_at":"2026-07-05T03:13:01.696930Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:13:01.696930Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"FastIF: Scalable Influence Functions for Efficient Model Interpretation and Debugging","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Caiming Xiong, Han Guo, Mohit Bansal, Nazneen Fatema Rajani, Peter Hase","submitted_at":"2020-12-31T18:02:34Z","abstract_excerpt":"Influence functions approximate the \"influences\" of training data-points for test predictions and have a wide variety of applications. Despite the popularity, their computational cost does not scale well with model and training data size. We present FastIF, a set of simple modifications to influence functions that significantly improves their run-time. We use k-Nearest Neighbors (kNN) to narrow the search space down to a subset of good candidate data points, identify the configurations that best balance the speed-quality trade-off in estimating the inverse Hessian-vector product, and introduce"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2012.15781","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/2012.15781/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":"2012.15781","created_at":"2026-07-05T03:13:01.696990+00:00"},{"alias_kind":"arxiv_version","alias_value":"2012.15781v2","created_at":"2026-07-05T03:13:01.696990+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2012.15781","created_at":"2026-07-05T03:13:01.696990+00:00"},{"alias_kind":"pith_short_12","alias_value":"HII5HMGUTHR6","created_at":"2026-07-05T03:13:01.696990+00:00"},{"alias_kind":"pith_short_16","alias_value":"HII5HMGUTHR6NY7D","created_at":"2026-07-05T03:13:01.696990+00:00"},{"alias_kind":"pith_short_8","alias_value":"HII5HMGU","created_at":"2026-07-05T03:13:01.696990+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.23223","citing_title":"Daunce: Data Attribution through Uncertainty Estimation","ref_index":4,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HII5HMGUTHR6NY7DP2C3HFGD2G","json":"https://pith.science/pith/HII5HMGUTHR6NY7DP2C3HFGD2G.json","graph_json":"https://pith.science/api/pith-number/HII5HMGUTHR6NY7DP2C3HFGD2G/graph.json","events_json":"https://pith.science/api/pith-number/HII5HMGUTHR6NY7DP2C3HFGD2G/events.json","paper":"https://pith.science/paper/HII5HMGU"},"agent_actions":{"view_html":"https://pith.science/pith/HII5HMGUTHR6NY7DP2C3HFGD2G","download_json":"https://pith.science/pith/HII5HMGUTHR6NY7DP2C3HFGD2G.json","view_paper":"https://pith.science/paper/HII5HMGU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2012.15781&json=true","fetch_graph":"https://pith.science/api/pith-number/HII5HMGUTHR6NY7DP2C3HFGD2G/graph.json","fetch_events":"https://pith.science/api/pith-number/HII5HMGUTHR6NY7DP2C3HFGD2G/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HII5HMGUTHR6NY7DP2C3HFGD2G/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HII5HMGUTHR6NY7DP2C3HFGD2G/action/storage_attestation","attest_author":"https://pith.science/pith/HII5HMGUTHR6NY7DP2C3HFGD2G/action/author_attestation","sign_citation":"https://pith.science/pith/HII5HMGUTHR6NY7DP2C3HFGD2G/action/citation_signature","submit_replication":"https://pith.science/pith/HII5HMGUTHR6NY7DP2C3HFGD2G/action/replication_record"}},"created_at":"2026-07-05T03:13:01.696990+00:00","updated_at":"2026-07-05T03:13:01.696990+00:00"}