{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:NG2OTSZIWLH66DAALWYDXLIUX2","short_pith_number":"pith:NG2OTSZI","schema_version":"1.0","canonical_sha256":"69b4e9cb28b2cfef0c005db03bad14beaeb5c9de436bdc1b2403789d6e9753ab","source":{"kind":"arxiv","id":"2106.10471","version":2},"attestation_state":"computed","paper":{"title":"Neural Network Classifier as Mutual Information Evaluator","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Dongwoo Kim, Tom Gedeon, Zhenyue Qin","submitted_at":"2021-06-19T10:55:46Z","abstract_excerpt":"Cross-entropy loss with softmax output is a standard choice to train neural network classifiers. We give a new view of neural network classifiers with softmax and cross-entropy as mutual information evaluators. We show that when the dataset is balanced, training a neural network with cross-entropy maximises the mutual information between inputs and labels through a variational form of mutual information. Thereby, we develop a new form of softmax that also converts a classifier to a mutual information evaluator when the dataset is imbalanced. Experimental results show that the new form leads to"},"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":"2106.10471","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-06-19T10:55:46Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"61c4272186b6cf8a03a9142fa74b67a6fa7186f719655c2621adac22c1784f3c","abstract_canon_sha256":"9b9216deeafeb81add47af5d67c85d970a03c2b594f94a009ee7df81220ff5b2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:05:51.351928Z","signature_b64":"3q67+lCWNDeBH5Da0i2z/SOtKimNMdLpBaoFPM77Tyd8iSNJxII231bZ6vTXee22Le7mfuRz3KjLpDeNOUm/Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"69b4e9cb28b2cfef0c005db03bad14beaeb5c9de436bdc1b2403789d6e9753ab","last_reissued_at":"2026-07-05T03:05:51.351414Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:05:51.351414Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Neural Network Classifier as Mutual Information Evaluator","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Dongwoo Kim, Tom Gedeon, Zhenyue Qin","submitted_at":"2021-06-19T10:55:46Z","abstract_excerpt":"Cross-entropy loss with softmax output is a standard choice to train neural network classifiers. We give a new view of neural network classifiers with softmax and cross-entropy as mutual information evaluators. We show that when the dataset is balanced, training a neural network with cross-entropy maximises the mutual information between inputs and labels through a variational form of mutual information. Thereby, we develop a new form of softmax that also converts a classifier to a mutual information evaluator when the dataset is imbalanced. Experimental results show that the new form leads to"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.10471","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/2106.10471/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":"2106.10471","created_at":"2026-07-05T03:05:51.351474+00:00"},{"alias_kind":"arxiv_version","alias_value":"2106.10471v2","created_at":"2026-07-05T03:05:51.351474+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.10471","created_at":"2026-07-05T03:05:51.351474+00:00"},{"alias_kind":"pith_short_12","alias_value":"NG2OTSZIWLH6","created_at":"2026-07-05T03:05:51.351474+00:00"},{"alias_kind":"pith_short_16","alias_value":"NG2OTSZIWLH66DAA","created_at":"2026-07-05T03:05:51.351474+00:00"},{"alias_kind":"pith_short_8","alias_value":"NG2OTSZI","created_at":"2026-07-05T03:05:51.351474+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/NG2OTSZIWLH66DAALWYDXLIUX2","json":"https://pith.science/pith/NG2OTSZIWLH66DAALWYDXLIUX2.json","graph_json":"https://pith.science/api/pith-number/NG2OTSZIWLH66DAALWYDXLIUX2/graph.json","events_json":"https://pith.science/api/pith-number/NG2OTSZIWLH66DAALWYDXLIUX2/events.json","paper":"https://pith.science/paper/NG2OTSZI"},"agent_actions":{"view_html":"https://pith.science/pith/NG2OTSZIWLH66DAALWYDXLIUX2","download_json":"https://pith.science/pith/NG2OTSZIWLH66DAALWYDXLIUX2.json","view_paper":"https://pith.science/paper/NG2OTSZI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2106.10471&json=true","fetch_graph":"https://pith.science/api/pith-number/NG2OTSZIWLH66DAALWYDXLIUX2/graph.json","fetch_events":"https://pith.science/api/pith-number/NG2OTSZIWLH66DAALWYDXLIUX2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NG2OTSZIWLH66DAALWYDXLIUX2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NG2OTSZIWLH66DAALWYDXLIUX2/action/storage_attestation","attest_author":"https://pith.science/pith/NG2OTSZIWLH66DAALWYDXLIUX2/action/author_attestation","sign_citation":"https://pith.science/pith/NG2OTSZIWLH66DAALWYDXLIUX2/action/citation_signature","submit_replication":"https://pith.science/pith/NG2OTSZIWLH66DAALWYDXLIUX2/action/replication_record"}},"created_at":"2026-07-05T03:05:51.351474+00:00","updated_at":"2026-07-05T03:05:51.351474+00:00"}