{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:XPXSDIBO6WSLJXRRH2RCOC6XVB","short_pith_number":"pith:XPXSDIBO","schema_version":"1.0","canonical_sha256":"bbef21a02ef5a4b4de313ea2270bd7a85573b09df1cdedcffc0a8ea6736cefa0","source":{"kind":"arxiv","id":"2007.07314","version":2},"attestation_state":"computed","paper":{"title":"Long-tail learning via logit adjustment","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Aditya Krishna Menon, Andreas Veit, Ankit Singh Rawat, Himanshu Jain, Sadeep Jayasumana, Sanjiv Kumar","submitted_at":"2020-07-14T19:27:13Z","abstract_excerpt":"Real-world classification problems typically exhibit an imbalanced or long-tailed label distribution, wherein many labels are associated with only a few samples. This poses a challenge for generalisation on such labels, and also makes na\\\"ive learning biased towards dominant labels. In this paper, we present two simple modifications of standard softmax cross-entropy training to cope with these challenges. Our techniques revisit the classic idea of logit adjustment based on the label frequencies, either applied post-hoc to a trained model, or enforced in the loss during training. Such adjustmen"},"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":"2007.07314","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2020-07-14T19:27:13Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"d881b3e030e7cd8a1bbe411db1549ba6365c3d7cee4d5eccc4567ab9c5fa4f0a","abstract_canon_sha256":"b081ec2c38b640d65cd4d50c723bce737972dd0f973f944fb55bfccf3621d2a1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:56:36.504381Z","signature_b64":"PmsfxBITnrjUNEQ7JaeJlYXwveVNuDW/DyqXHujy1w+NQhtKsL7BCPZcIYqYb/1nkMyAWSh0ez0OTs/o6TTFBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bbef21a02ef5a4b4de313ea2270bd7a85573b09df1cdedcffc0a8ea6736cefa0","last_reissued_at":"2026-07-05T02:56:36.503900Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:56:36.503900Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Long-tail learning via logit adjustment","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Aditya Krishna Menon, Andreas Veit, Ankit Singh Rawat, Himanshu Jain, Sadeep Jayasumana, Sanjiv Kumar","submitted_at":"2020-07-14T19:27:13Z","abstract_excerpt":"Real-world classification problems typically exhibit an imbalanced or long-tailed label distribution, wherein many labels are associated with only a few samples. This poses a challenge for generalisation on such labels, and also makes na\\\"ive learning biased towards dominant labels. In this paper, we present two simple modifications of standard softmax cross-entropy training to cope with these challenges. Our techniques revisit the classic idea of logit adjustment based on the label frequencies, either applied post-hoc to a trained model, or enforced in the loss during training. Such adjustmen"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2007.07314","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/2007.07314/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":"2007.07314","created_at":"2026-07-05T02:56:36.503976+00:00"},{"alias_kind":"arxiv_version","alias_value":"2007.07314v2","created_at":"2026-07-05T02:56:36.503976+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2007.07314","created_at":"2026-07-05T02:56:36.503976+00:00"},{"alias_kind":"pith_short_12","alias_value":"XPXSDIBO6WSL","created_at":"2026-07-05T02:56:36.503976+00:00"},{"alias_kind":"pith_short_16","alias_value":"XPXSDIBO6WSLJXRR","created_at":"2026-07-05T02:56:36.503976+00:00"},{"alias_kind":"pith_short_8","alias_value":"XPXSDIBO","created_at":"2026-07-05T02:56:36.503976+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":14,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.06176","citing_title":"Revisiting Scene Graph Generation from the Perspective of Detector-Conditioned Reachability","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"2606.22702","citing_title":"Modular Diffusion Models for Structured Visual Recognition","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00057","citing_title":"Enhancing Oracle Bone Inscription Recognition via Multi-Scale Layer Attention","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04363","citing_title":"Mitigating Label Shift in Tabular In-Context Learning via Test-Time Posterior Adjustment","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2606.27696","citing_title":"Class-frequency Guided Noise Schedule for Diffusion Models","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2408.11338","citing_title":"Automatic Dataset Construction (ADC): Sample Collection, Data Curation, and Beyond","ref_index":76,"is_internal_anchor":false},{"citing_arxiv_id":"2511.04838","citing_title":"SPECTRA: Spectral Domain-Aware Graph Generation for Imbalanced Molecular Property Regression","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2509.09926","citing_title":"LoFT: Parameter-Efficient Fine-Tuning for Long-tailed Semi-Supervised Learning in Open-World Scenarios","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2512.18994","citing_title":"Dual-Margin Embedding for Fine-Grained Long-Tailed Plant Taxonomy","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03364","citing_title":"Dynamic Distillation and Gradient Consistency for Robust Long-Tailed Incremental Learning","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10498","citing_title":"Simultaneous Long-tailed Recognition and Multi-modal Fusion for Highly Imbalanced Multi-modal Data","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10047","citing_title":"Rethinking Loss Reweighting for Imbalance Learning as an Inverse Problem: A Neural Collapse Point of View","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09737","citing_title":"STaR-DRO: Stateful Tsallis Reweighting for Group-Robust Structured Prediction","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04363","citing_title":"Mitigating Label Shift in Tabular In-Context Learning via Test-Time Posterior Adjustment","ref_index":6,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XPXSDIBO6WSLJXRRH2RCOC6XVB","json":"https://pith.science/pith/XPXSDIBO6WSLJXRRH2RCOC6XVB.json","graph_json":"https://pith.science/api/pith-number/XPXSDIBO6WSLJXRRH2RCOC6XVB/graph.json","events_json":"https://pith.science/api/pith-number/XPXSDIBO6WSLJXRRH2RCOC6XVB/events.json","paper":"https://pith.science/paper/XPXSDIBO"},"agent_actions":{"view_html":"https://pith.science/pith/XPXSDIBO6WSLJXRRH2RCOC6XVB","download_json":"https://pith.science/pith/XPXSDIBO6WSLJXRRH2RCOC6XVB.json","view_paper":"https://pith.science/paper/XPXSDIBO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2007.07314&json=true","fetch_graph":"https://pith.science/api/pith-number/XPXSDIBO6WSLJXRRH2RCOC6XVB/graph.json","fetch_events":"https://pith.science/api/pith-number/XPXSDIBO6WSLJXRRH2RCOC6XVB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XPXSDIBO6WSLJXRRH2RCOC6XVB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XPXSDIBO6WSLJXRRH2RCOC6XVB/action/storage_attestation","attest_author":"https://pith.science/pith/XPXSDIBO6WSLJXRRH2RCOC6XVB/action/author_attestation","sign_citation":"https://pith.science/pith/XPXSDIBO6WSLJXRRH2RCOC6XVB/action/citation_signature","submit_replication":"https://pith.science/pith/XPXSDIBO6WSLJXRRH2RCOC6XVB/action/replication_record"}},"created_at":"2026-07-05T02:56:36.503976+00:00","updated_at":"2026-07-05T02:56:36.503976+00:00"}