{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:JTANOIS5RWW3T6QWTBEBREE6WO","short_pith_number":"pith:JTANOIS5","schema_version":"1.0","canonical_sha256":"4cc0d7225d8dadb9fa16984818909eb3b74fded211e92cb9665de4f478df3648","source":{"kind":"arxiv","id":"2210.07126","version":2},"attestation_state":"computed","paper":{"title":"Challenges in Explanation Quality Evaluation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.HC"],"primary_cat":"cs.CL","authors_text":"Heike Adel, Hendrik Schuff, Ngoc Thang Vu, Peng Qi","submitted_at":"2022-10-13T16:06:59Z","abstract_excerpt":"While much research focused on producing explanations, it is still unclear how the produced explanations' quality can be evaluated in a meaningful way. Today's predominant approach is to quantify explanations using proxy scores which compare explanations to (human-annotated) gold explanations. This approach assumes that explanations which reach higher proxy scores will also provide a greater benefit to human users. In this paper, we present problems of this approach. Concretely, we (i) formulate desired characteristics of explanation quality, (ii) describe how current evaluation practices viol"},"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":"2210.07126","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-10-13T16:06:59Z","cross_cats_sorted":["cs.AI","cs.HC"],"title_canon_sha256":"1b2e0f8011b45791f8bac024575d992b87f52b04a5a63cf6b61006cab09fd8bd","abstract_canon_sha256":"b2f9de97f4b284a654d5de0f2458f22541b3bb17adf1e7685868c3b18068ecfb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:49:30.675601Z","signature_b64":"uCm+yZthi3A4GVjvyIZqUZFGpCjrMjVns3U5WvVDZhzfirMo1+kgpCXUEeH7q9YH2SbVqJ1gIygOXVkOXXPQDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4cc0d7225d8dadb9fa16984818909eb3b74fded211e92cb9665de4f478df3648","last_reissued_at":"2026-07-05T05:49:30.675171Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:49:30.675171Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Challenges in Explanation Quality Evaluation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.HC"],"primary_cat":"cs.CL","authors_text":"Heike Adel, Hendrik Schuff, Ngoc Thang Vu, Peng Qi","submitted_at":"2022-10-13T16:06:59Z","abstract_excerpt":"While much research focused on producing explanations, it is still unclear how the produced explanations' quality can be evaluated in a meaningful way. Today's predominant approach is to quantify explanations using proxy scores which compare explanations to (human-annotated) gold explanations. This approach assumes that explanations which reach higher proxy scores will also provide a greater benefit to human users. In this paper, we present problems of this approach. Concretely, we (i) formulate desired characteristics of explanation quality, (ii) describe how current evaluation practices viol"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.07126","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/2210.07126/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":"2210.07126","created_at":"2026-07-05T05:49:30.675231+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.07126v2","created_at":"2026-07-05T05:49:30.675231+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.07126","created_at":"2026-07-05T05:49:30.675231+00:00"},{"alias_kind":"pith_short_12","alias_value":"JTANOIS5RWW3","created_at":"2026-07-05T05:49:30.675231+00:00"},{"alias_kind":"pith_short_16","alias_value":"JTANOIS5RWW3T6QW","created_at":"2026-07-05T05:49:30.675231+00:00"},{"alias_kind":"pith_short_8","alias_value":"JTANOIS5","created_at":"2026-07-05T05:49:30.675231+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.05790","citing_title":"Improving Explanations: Applying the Feature Understandability Scale for Cost-Sensitive Feature Selection","ref_index":51,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JTANOIS5RWW3T6QWTBEBREE6WO","json":"https://pith.science/pith/JTANOIS5RWW3T6QWTBEBREE6WO.json","graph_json":"https://pith.science/api/pith-number/JTANOIS5RWW3T6QWTBEBREE6WO/graph.json","events_json":"https://pith.science/api/pith-number/JTANOIS5RWW3T6QWTBEBREE6WO/events.json","paper":"https://pith.science/paper/JTANOIS5"},"agent_actions":{"view_html":"https://pith.science/pith/JTANOIS5RWW3T6QWTBEBREE6WO","download_json":"https://pith.science/pith/JTANOIS5RWW3T6QWTBEBREE6WO.json","view_paper":"https://pith.science/paper/JTANOIS5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.07126&json=true","fetch_graph":"https://pith.science/api/pith-number/JTANOIS5RWW3T6QWTBEBREE6WO/graph.json","fetch_events":"https://pith.science/api/pith-number/JTANOIS5RWW3T6QWTBEBREE6WO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JTANOIS5RWW3T6QWTBEBREE6WO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JTANOIS5RWW3T6QWTBEBREE6WO/action/storage_attestation","attest_author":"https://pith.science/pith/JTANOIS5RWW3T6QWTBEBREE6WO/action/author_attestation","sign_citation":"https://pith.science/pith/JTANOIS5RWW3T6QWTBEBREE6WO/action/citation_signature","submit_replication":"https://pith.science/pith/JTANOIS5RWW3T6QWTBEBREE6WO/action/replication_record"}},"created_at":"2026-07-05T05:49:30.675231+00:00","updated_at":"2026-07-05T05:49:30.675231+00:00"}