{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:WP3JYO2GSLWXBWH7E5KUKITLHZ","short_pith_number":"pith:WP3JYO2G","schema_version":"1.0","canonical_sha256":"b3f69c3b4692ed70d8ff275545226b3e4f5d9740974695f66d7801bbc9157d59","source":{"kind":"arxiv","id":"2206.07290","version":1},"attestation_state":"computed","paper":{"title":"Differentiable Top-k Classification Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Christian Borgelt, Felix Petersen, Hilde Kuehne, Oliver Deussen","submitted_at":"2022-06-15T04:13:59Z","abstract_excerpt":"The top-k classification accuracy is one of the core metrics in machine learning. Here, k is conventionally a positive integer, such as 1 or 5, leading to top-1 or top-5 training objectives. In this work, we relax this assumption and optimize the model for multiple k simultaneously instead of using a single k. Leveraging recent advances in differentiable sorting and ranking, we propose a differentiable top-k cross-entropy classification loss. This allows training the network while not only considering the top-1 prediction, but also, e.g., the top-2 and top-5 predictions. We evaluate the propos"},"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":"2206.07290","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-06-15T04:13:59Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"3ce6f6ccf40326323b87a697a3a6d7b38e93eef3225a2b9f8d090a8b15d1b1fc","abstract_canon_sha256":"4c6d04f836146efba2be776b1c82ceeebe11b355cccce84c06f3953b44dfd852"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:32:03.095368Z","signature_b64":"rXAa272Zv89itQdMO2Q8JQubX1j2GJSh7CagjtdQL77nPgKgRw1bOV1jl6E4ED9+gvmzi4Qoe6REc10VBdCWAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b3f69c3b4692ed70d8ff275545226b3e4f5d9740974695f66d7801bbc9157d59","last_reissued_at":"2026-07-05T04:32:03.094853Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:32:03.094853Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Differentiable Top-k Classification Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Christian Borgelt, Felix Petersen, Hilde Kuehne, Oliver Deussen","submitted_at":"2022-06-15T04:13:59Z","abstract_excerpt":"The top-k classification accuracy is one of the core metrics in machine learning. Here, k is conventionally a positive integer, such as 1 or 5, leading to top-1 or top-5 training objectives. In this work, we relax this assumption and optimize the model for multiple k simultaneously instead of using a single k. Leveraging recent advances in differentiable sorting and ranking, we propose a differentiable top-k cross-entropy classification loss. This allows training the network while not only considering the top-1 prediction, but also, e.g., the top-2 and top-5 predictions. We evaluate the propos"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2206.07290","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/2206.07290/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":"2206.07290","created_at":"2026-07-05T04:32:03.094920+00:00"},{"alias_kind":"arxiv_version","alias_value":"2206.07290v1","created_at":"2026-07-05T04:32:03.094920+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2206.07290","created_at":"2026-07-05T04:32:03.094920+00:00"},{"alias_kind":"pith_short_12","alias_value":"WP3JYO2GSLWX","created_at":"2026-07-05T04:32:03.094920+00:00"},{"alias_kind":"pith_short_16","alias_value":"WP3JYO2GSLWXBWH7","created_at":"2026-07-05T04:32:03.094920+00:00"},{"alias_kind":"pith_short_8","alias_value":"WP3JYO2G","created_at":"2026-07-05T04:32:03.094920+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.12655","citing_title":"Robust Instruction Compliance in Cooperative Multi-Agent Reinforcement Learning","ref_index":176,"is_internal_anchor":false},{"citing_arxiv_id":"2605.25551","citing_title":"Learning Permutation from Structure Without Supervision","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WP3JYO2GSLWXBWH7E5KUKITLHZ","json":"https://pith.science/pith/WP3JYO2GSLWXBWH7E5KUKITLHZ.json","graph_json":"https://pith.science/api/pith-number/WP3JYO2GSLWXBWH7E5KUKITLHZ/graph.json","events_json":"https://pith.science/api/pith-number/WP3JYO2GSLWXBWH7E5KUKITLHZ/events.json","paper":"https://pith.science/paper/WP3JYO2G"},"agent_actions":{"view_html":"https://pith.science/pith/WP3JYO2GSLWXBWH7E5KUKITLHZ","download_json":"https://pith.science/pith/WP3JYO2GSLWXBWH7E5KUKITLHZ.json","view_paper":"https://pith.science/paper/WP3JYO2G","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2206.07290&json=true","fetch_graph":"https://pith.science/api/pith-number/WP3JYO2GSLWXBWH7E5KUKITLHZ/graph.json","fetch_events":"https://pith.science/api/pith-number/WP3JYO2GSLWXBWH7E5KUKITLHZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WP3JYO2GSLWXBWH7E5KUKITLHZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WP3JYO2GSLWXBWH7E5KUKITLHZ/action/storage_attestation","attest_author":"https://pith.science/pith/WP3JYO2GSLWXBWH7E5KUKITLHZ/action/author_attestation","sign_citation":"https://pith.science/pith/WP3JYO2GSLWXBWH7E5KUKITLHZ/action/citation_signature","submit_replication":"https://pith.science/pith/WP3JYO2GSLWXBWH7E5KUKITLHZ/action/replication_record"}},"created_at":"2026-07-05T04:32:03.094920+00:00","updated_at":"2026-07-05T04:32:03.094920+00:00"}