{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:RH6MIZNTY3H4XIHB76QTNTW53N","short_pith_number":"pith:RH6MIZNT","schema_version":"1.0","canonical_sha256":"89fcc465b3c6cfcba0e1ffa136cedddb5ebbad4629ec2e6c58ed50ef1b92eb6c","source":{"kind":"arxiv","id":"2312.04893","version":1},"attestation_state":"computed","paper":{"title":"Annotation-Free Group Robustness via Loss-Based Resampling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Fahimeh Hosseini Noohdani, HamidReza Yaghoubi Araghi, Hesam Asadollahzadeh, Mahdieh Soleymani Baghshah, Mahdi Ghaznavi, Mohammad Hossein Rohban","submitted_at":"2023-12-08T08:22:02Z","abstract_excerpt":"It is well-known that training neural networks for image classification with empirical risk minimization (ERM) makes them vulnerable to relying on spurious attributes instead of causal ones for prediction. Previously, deep feature re-weighting (DFR) has proposed retraining the last layer of a pre-trained network on balanced data concerning spurious attributes, making it robust to spurious correlation. However, spurious attribute annotations are not always available. In order to provide group robustness without such annotations, we propose a new method, called loss-based feature re-weighting (L"},"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":"2312.04893","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-12-08T08:22:02Z","cross_cats_sorted":[],"title_canon_sha256":"c98cf4b2f747ab6247ea8b209526d11f7b865b95792fd924dc14492e4896c557","abstract_canon_sha256":"d1157c10434baf9ed215daaa0f8248f10130f55a15df9cdb9edb808247ea2ee3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:21:53.900692Z","signature_b64":"UOGwhSjQ0/NxlWAF13GJoLVmAZ91Y6T8VWVBXsROLqZ9HsgBxkN7E5m3lUqoHrGeZ7NjwTUg3xiLo/WVhqrPDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"89fcc465b3c6cfcba0e1ffa136cedddb5ebbad4629ec2e6c58ed50ef1b92eb6c","last_reissued_at":"2026-07-05T07:21:53.900176Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:21:53.900176Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Annotation-Free Group Robustness via Loss-Based Resampling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Fahimeh Hosseini Noohdani, HamidReza Yaghoubi Araghi, Hesam Asadollahzadeh, Mahdieh Soleymani Baghshah, Mahdi Ghaznavi, Mohammad Hossein Rohban","submitted_at":"2023-12-08T08:22:02Z","abstract_excerpt":"It is well-known that training neural networks for image classification with empirical risk minimization (ERM) makes them vulnerable to relying on spurious attributes instead of causal ones for prediction. Previously, deep feature re-weighting (DFR) has proposed retraining the last layer of a pre-trained network on balanced data concerning spurious attributes, making it robust to spurious correlation. However, spurious attribute annotations are not always available. In order to provide group robustness without such annotations, we propose a new method, called loss-based feature re-weighting (L"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.04893","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/2312.04893/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":"2312.04893","created_at":"2026-07-05T07:21:53.900228+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.04893v1","created_at":"2026-07-05T07:21:53.900228+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.04893","created_at":"2026-07-05T07:21:53.900228+00:00"},{"alias_kind":"pith_short_12","alias_value":"RH6MIZNTY3H4","created_at":"2026-07-05T07:21:53.900228+00:00"},{"alias_kind":"pith_short_16","alias_value":"RH6MIZNTY3H4XIHB","created_at":"2026-07-05T07:21:53.900228+00:00"},{"alias_kind":"pith_short_8","alias_value":"RH6MIZNT","created_at":"2026-07-05T07:21:53.900228+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/RH6MIZNTY3H4XIHB76QTNTW53N","json":"https://pith.science/pith/RH6MIZNTY3H4XIHB76QTNTW53N.json","graph_json":"https://pith.science/api/pith-number/RH6MIZNTY3H4XIHB76QTNTW53N/graph.json","events_json":"https://pith.science/api/pith-number/RH6MIZNTY3H4XIHB76QTNTW53N/events.json","paper":"https://pith.science/paper/RH6MIZNT"},"agent_actions":{"view_html":"https://pith.science/pith/RH6MIZNTY3H4XIHB76QTNTW53N","download_json":"https://pith.science/pith/RH6MIZNTY3H4XIHB76QTNTW53N.json","view_paper":"https://pith.science/paper/RH6MIZNT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.04893&json=true","fetch_graph":"https://pith.science/api/pith-number/RH6MIZNTY3H4XIHB76QTNTW53N/graph.json","fetch_events":"https://pith.science/api/pith-number/RH6MIZNTY3H4XIHB76QTNTW53N/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RH6MIZNTY3H4XIHB76QTNTW53N/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RH6MIZNTY3H4XIHB76QTNTW53N/action/storage_attestation","attest_author":"https://pith.science/pith/RH6MIZNTY3H4XIHB76QTNTW53N/action/author_attestation","sign_citation":"https://pith.science/pith/RH6MIZNTY3H4XIHB76QTNTW53N/action/citation_signature","submit_replication":"https://pith.science/pith/RH6MIZNTY3H4XIHB76QTNTW53N/action/replication_record"}},"created_at":"2026-07-05T07:21:53.900228+00:00","updated_at":"2026-07-05T07:21:53.900228+00:00"}