{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:VZDMXS5NUDQB6UCMAENBR5UDKY","short_pith_number":"pith:VZDMXS5N","schema_version":"1.0","canonical_sha256":"ae46cbcbada0e01f504c011a18f68356238a0b71169d93c822934678de4561d2","source":{"kind":"arxiv","id":"1910.08485","version":1},"attestation_state":"computed","paper":{"title":"Understanding Deep Networks via Extremal Perturbations and Smooth Masks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CV","authors_text":"Andrea Vedaldi, Mandela Patrick, Ruth Fong","submitted_at":"2019-10-18T16:02:01Z","abstract_excerpt":"The problem of attribution is concerned with identifying the parts of an input that are responsible for a model's output. An important family of attribution methods is based on measuring the effect of perturbations applied to the input. In this paper, we discuss some of the shortcomings of existing approaches to perturbation analysis and address them by introducing the concept of extremal perturbations, which are theoretically grounded and interpretable. We also introduce a number of technical innovations to compute extremal perturbations, including a new area constraint and a parametric famil"},"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":"1910.08485","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2019-10-18T16:02:01Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"347546b408d0112e1ea4aec944f041a11965db09e4e442a8e5b7f8d9bd29d77f","abstract_canon_sha256":"802cf2e71204f884f588cb6b03d696380503870f6925db69459956660db50f15"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:13:09.539503Z","signature_b64":"5qkp0JmAG3Htr7MAPaM6d6a7iPMqyGRVDJRHoHg4lOzgAvxuC+MXR4S2gDCU1uksHqGVcny9IWVYz3sIOtpjAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ae46cbcbada0e01f504c011a18f68356238a0b71169d93c822934678de4561d2","last_reissued_at":"2026-07-05T00:13:09.539103Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:13:09.539103Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Understanding Deep Networks via Extremal Perturbations and Smooth Masks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CV","authors_text":"Andrea Vedaldi, Mandela Patrick, Ruth Fong","submitted_at":"2019-10-18T16:02:01Z","abstract_excerpt":"The problem of attribution is concerned with identifying the parts of an input that are responsible for a model's output. An important family of attribution methods is based on measuring the effect of perturbations applied to the input. In this paper, we discuss some of the shortcomings of existing approaches to perturbation analysis and address them by introducing the concept of extremal perturbations, which are theoretically grounded and interpretable. We also introduce a number of technical innovations to compute extremal perturbations, including a new area constraint and a parametric famil"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1910.08485","kind":"arxiv","version":1},"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/1910.08485/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":"1910.08485","created_at":"2026-07-05T00:13:09.539158+00:00"},{"alias_kind":"arxiv_version","alias_value":"1910.08485v1","created_at":"2026-07-05T00:13:09.539158+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1910.08485","created_at":"2026-07-05T00:13:09.539158+00:00"},{"alias_kind":"pith_short_12","alias_value":"VZDMXS5NUDQB","created_at":"2026-07-05T00:13:09.539158+00:00"},{"alias_kind":"pith_short_16","alias_value":"VZDMXS5NUDQB6UCM","created_at":"2026-07-05T00:13:09.539158+00:00"},{"alias_kind":"pith_short_8","alias_value":"VZDMXS5N","created_at":"2026-07-05T00:13:09.539158+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.27618","citing_title":"Evaluating Local Explainability Metrics for Machine Learning Models on Tabular Data","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2511.01411","citing_title":"Extremal Contours: Gradient-driven contours for compact visual attribution","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VZDMXS5NUDQB6UCMAENBR5UDKY","json":"https://pith.science/pith/VZDMXS5NUDQB6UCMAENBR5UDKY.json","graph_json":"https://pith.science/api/pith-number/VZDMXS5NUDQB6UCMAENBR5UDKY/graph.json","events_json":"https://pith.science/api/pith-number/VZDMXS5NUDQB6UCMAENBR5UDKY/events.json","paper":"https://pith.science/paper/VZDMXS5N"},"agent_actions":{"view_html":"https://pith.science/pith/VZDMXS5NUDQB6UCMAENBR5UDKY","download_json":"https://pith.science/pith/VZDMXS5NUDQB6UCMAENBR5UDKY.json","view_paper":"https://pith.science/paper/VZDMXS5N","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1910.08485&json=true","fetch_graph":"https://pith.science/api/pith-number/VZDMXS5NUDQB6UCMAENBR5UDKY/graph.json","fetch_events":"https://pith.science/api/pith-number/VZDMXS5NUDQB6UCMAENBR5UDKY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VZDMXS5NUDQB6UCMAENBR5UDKY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VZDMXS5NUDQB6UCMAENBR5UDKY/action/storage_attestation","attest_author":"https://pith.science/pith/VZDMXS5NUDQB6UCMAENBR5UDKY/action/author_attestation","sign_citation":"https://pith.science/pith/VZDMXS5NUDQB6UCMAENBR5UDKY/action/citation_signature","submit_replication":"https://pith.science/pith/VZDMXS5NUDQB6UCMAENBR5UDKY/action/replication_record"}},"created_at":"2026-07-05T00:13:09.539158+00:00","updated_at":"2026-07-05T00:13:09.539158+00:00"}