{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:NTSFHQPI274M3ETXKUMXMDXIHA","short_pith_number":"pith:NTSFHQPI","schema_version":"1.0","canonical_sha256":"6ce453c1e8d7f8cd92775519760ee83809066a3b80bf5dcef9f7c072ba0a57ab","source":{"kind":"arxiv","id":"1910.09086","version":2},"attestation_state":"computed","paper":{"title":"Contextual Prediction Difference Analysis for Explaining Individual Image Classifications","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jindong Gu, Volker Tresp","submitted_at":"2019-10-21T00:04:22Z","abstract_excerpt":"Much effort has been devoted to understanding the decisions of deep neural networks in recent years. A number of model-aware saliency methods were proposed to explain individual classification decisions by creating saliency maps. However, they are not applicable when the parameters and the gradients of the underlying models are unavailable. Recently, model-agnostic methods have also received attention. As one of them, \\textit{Prediction Difference Analysis} (PDA), a probabilistic sound methodology, was proposed. In this work, we first show that PDA can suffer from saturated classifiers. The sa"},"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":"1910.09086","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-10-21T00:04:22Z","cross_cats_sorted":["cs.CV","stat.ML"],"title_canon_sha256":"a74f1169ddc7589937be2c97db01adba937ae9c641d679c378603a31dd55d344","abstract_canon_sha256":"c7bd32b477b017e7990404bd339b3fe989b21f6344e3e4e4fadae2c068caad8b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:08:25.223616Z","signature_b64":"zTHRDgQI+TQCfwOCjENBzS/OdLMyrnHsp4HoAMJFY+NWk5r5NOEmKuJUZKSiB40PWDOWZAZUZVrNz4JT/WkECQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6ce453c1e8d7f8cd92775519760ee83809066a3b80bf5dcef9f7c072ba0a57ab","last_reissued_at":"2026-07-05T01:08:25.223245Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:08:25.223245Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Contextual Prediction Difference Analysis for Explaining Individual Image Classifications","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jindong Gu, Volker Tresp","submitted_at":"2019-10-21T00:04:22Z","abstract_excerpt":"Much effort has been devoted to understanding the decisions of deep neural networks in recent years. A number of model-aware saliency methods were proposed to explain individual classification decisions by creating saliency maps. However, they are not applicable when the parameters and the gradients of the underlying models are unavailable. Recently, model-agnostic methods have also received attention. As one of them, \\textit{Prediction Difference Analysis} (PDA), a probabilistic sound methodology, was proposed. In this work, we first show that PDA can suffer from saturated classifiers. The sa"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1910.09086","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/1910.09086/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":"1910.09086","created_at":"2026-07-05T01:08:25.223308+00:00"},{"alias_kind":"arxiv_version","alias_value":"1910.09086v2","created_at":"2026-07-05T01:08:25.223308+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1910.09086","created_at":"2026-07-05T01:08:25.223308+00:00"},{"alias_kind":"pith_short_12","alias_value":"NTSFHQPI274M","created_at":"2026-07-05T01:08:25.223308+00:00"},{"alias_kind":"pith_short_16","alias_value":"NTSFHQPI274M3ETX","created_at":"2026-07-05T01:08:25.223308+00:00"},{"alias_kind":"pith_short_8","alias_value":"NTSFHQPI","created_at":"2026-07-05T01:08:25.223308+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/NTSFHQPI274M3ETXKUMXMDXIHA","json":"https://pith.science/pith/NTSFHQPI274M3ETXKUMXMDXIHA.json","graph_json":"https://pith.science/api/pith-number/NTSFHQPI274M3ETXKUMXMDXIHA/graph.json","events_json":"https://pith.science/api/pith-number/NTSFHQPI274M3ETXKUMXMDXIHA/events.json","paper":"https://pith.science/paper/NTSFHQPI"},"agent_actions":{"view_html":"https://pith.science/pith/NTSFHQPI274M3ETXKUMXMDXIHA","download_json":"https://pith.science/pith/NTSFHQPI274M3ETXKUMXMDXIHA.json","view_paper":"https://pith.science/paper/NTSFHQPI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1910.09086&json=true","fetch_graph":"https://pith.science/api/pith-number/NTSFHQPI274M3ETXKUMXMDXIHA/graph.json","fetch_events":"https://pith.science/api/pith-number/NTSFHQPI274M3ETXKUMXMDXIHA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NTSFHQPI274M3ETXKUMXMDXIHA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NTSFHQPI274M3ETXKUMXMDXIHA/action/storage_attestation","attest_author":"https://pith.science/pith/NTSFHQPI274M3ETXKUMXMDXIHA/action/author_attestation","sign_citation":"https://pith.science/pith/NTSFHQPI274M3ETXKUMXMDXIHA/action/citation_signature","submit_replication":"https://pith.science/pith/NTSFHQPI274M3ETXKUMXMDXIHA/action/replication_record"}},"created_at":"2026-07-05T01:08:25.223308+00:00","updated_at":"2026-07-05T01:08:25.223308+00:00"}