{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:BQJYBPVNQE4W5ILNOG2FIFCAY7","short_pith_number":"pith:BQJYBPVN","schema_version":"1.0","canonical_sha256":"0c1380bead81396ea16d71b4541440c7f366d45dc41e0e3199ac224036d015bb","source":{"kind":"arxiv","id":"2104.14403","version":2},"attestation_state":"computed","paper":{"title":"Do Feature Attribution Methods Correctly Attribute Features?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Julie Shah, Marco Tulio Ribeiro, Serena Booth, Yilun Zhou","submitted_at":"2021-04-27T20:35:30Z","abstract_excerpt":"Feature attribution methods are popular in interpretable machine learning. These methods compute the attribution of each input feature to represent its importance, but there is no consensus on the definition of \"attribution\", leading to many competing methods with little systematic evaluation, complicated in particular by the lack of ground truth attribution. To address this, we propose a dataset modification procedure to induce such ground truth. Using this procedure, we evaluate three common methods: saliency maps, rationales, and attentions. We identify several deficiencies and add new pers"},"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":"2104.14403","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-04-27T20:35:30Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"fce646764904b5ecd4591f2cc89e68044d79a18e902056258aae6e1421afec91","abstract_canon_sha256":"45f66656a8570cedcba6871339befd3dc1e8b36f05554aba1f668a685f5b3e41"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:41:16.119999Z","signature_b64":"QRkKGr+1w7fufAYuEK18KuDWe+livUxY9nLT5dUtkvQHDFrlmI5iXcjSsid4cZoPcRbWF6S58s+LCG88hWqTBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0c1380bead81396ea16d71b4541440c7f366d45dc41e0e3199ac224036d015bb","last_reissued_at":"2026-07-05T03:41:16.119602Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:41:16.119602Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Do Feature Attribution Methods Correctly Attribute Features?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Julie Shah, Marco Tulio Ribeiro, Serena Booth, Yilun Zhou","submitted_at":"2021-04-27T20:35:30Z","abstract_excerpt":"Feature attribution methods are popular in interpretable machine learning. These methods compute the attribution of each input feature to represent its importance, but there is no consensus on the definition of \"attribution\", leading to many competing methods with little systematic evaluation, complicated in particular by the lack of ground truth attribution. To address this, we propose a dataset modification procedure to induce such ground truth. Using this procedure, we evaluate three common methods: saliency maps, rationales, and attentions. We identify several deficiencies and add new pers"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2104.14403","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/2104.14403/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":"2104.14403","created_at":"2026-07-05T03:41:16.119658+00:00"},{"alias_kind":"arxiv_version","alias_value":"2104.14403v2","created_at":"2026-07-05T03:41:16.119658+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2104.14403","created_at":"2026-07-05T03:41:16.119658+00:00"},{"alias_kind":"pith_short_12","alias_value":"BQJYBPVNQE4W","created_at":"2026-07-05T03:41:16.119658+00:00"},{"alias_kind":"pith_short_16","alias_value":"BQJYBPVNQE4W5ILN","created_at":"2026-07-05T03:41:16.119658+00:00"},{"alias_kind":"pith_short_8","alias_value":"BQJYBPVN","created_at":"2026-07-05T03:41:16.119658+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/BQJYBPVNQE4W5ILNOG2FIFCAY7","json":"https://pith.science/pith/BQJYBPVNQE4W5ILNOG2FIFCAY7.json","graph_json":"https://pith.science/api/pith-number/BQJYBPVNQE4W5ILNOG2FIFCAY7/graph.json","events_json":"https://pith.science/api/pith-number/BQJYBPVNQE4W5ILNOG2FIFCAY7/events.json","paper":"https://pith.science/paper/BQJYBPVN"},"agent_actions":{"view_html":"https://pith.science/pith/BQJYBPVNQE4W5ILNOG2FIFCAY7","download_json":"https://pith.science/pith/BQJYBPVNQE4W5ILNOG2FIFCAY7.json","view_paper":"https://pith.science/paper/BQJYBPVN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2104.14403&json=true","fetch_graph":"https://pith.science/api/pith-number/BQJYBPVNQE4W5ILNOG2FIFCAY7/graph.json","fetch_events":"https://pith.science/api/pith-number/BQJYBPVNQE4W5ILNOG2FIFCAY7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BQJYBPVNQE4W5ILNOG2FIFCAY7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BQJYBPVNQE4W5ILNOG2FIFCAY7/action/storage_attestation","attest_author":"https://pith.science/pith/BQJYBPVNQE4W5ILNOG2FIFCAY7/action/author_attestation","sign_citation":"https://pith.science/pith/BQJYBPVNQE4W5ILNOG2FIFCAY7/action/citation_signature","submit_replication":"https://pith.science/pith/BQJYBPVNQE4W5ILNOG2FIFCAY7/action/replication_record"}},"created_at":"2026-07-05T03:41:16.119658+00:00","updated_at":"2026-07-05T03:41:16.119658+00:00"}