{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:VJTCAPDK56PU4XHCMSKUSIVQNL","short_pith_number":"pith:VJTCAPDK","schema_version":"1.0","canonical_sha256":"aa66203c6aef9f4e5ce264954922b06ad09386b967238c3b1372c3eb67a9c6b4","source":{"kind":"arxiv","id":"2011.07932","version":4},"attestation_state":"computed","paper":{"title":"Combating the Instability of Mutual Information-based Losses via Regularization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IT","math.IT","stat.ML"],"primary_cat":"cs.LG","authors_text":"Kwanghee Choi, Siyeong Lee","submitted_at":"2020-11-16T13:29:15Z","abstract_excerpt":"Notable progress has been made in numerous fields of machine learning based on neural network-driven mutual information (MI) bounds. However, utilizing the conventional MI-based losses is often challenging due to their practical and mathematical limitations. In this work, we first identify the symptoms behind their instability: (1) the neural network not converging even after the loss seemed to converge, and (2) saturating neural network outputs causing the loss to diverge. We mitigate both issues by adding a novel regularization term to the existing losses. We theoretically and experimentally"},"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":"2011.07932","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-11-16T13:29:15Z","cross_cats_sorted":["cs.IT","math.IT","stat.ML"],"title_canon_sha256":"a95b7053d551b4f4605b42cb092080adf938c0a010cc25e60365190898483afa","abstract_canon_sha256":"934baf3c1ccacc4381d0a9ba6fd930c654c90602bc59a729910fe87bd549ad2a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:32:48.285280Z","signature_b64":"Z6HgoAeAPWpKS3zTkM8lur11m1r1g9wjSMvoHVpxEXDvJmQ6JE0Nozi0tCfgYuuSK3Qx2cTZOKGrLUAdB3cUCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aa66203c6aef9f4e5ce264954922b06ad09386b967238c3b1372c3eb67a9c6b4","last_reissued_at":"2026-07-05T04:32:48.284851Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:32:48.284851Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Combating the Instability of Mutual Information-based Losses via Regularization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IT","math.IT","stat.ML"],"primary_cat":"cs.LG","authors_text":"Kwanghee Choi, Siyeong Lee","submitted_at":"2020-11-16T13:29:15Z","abstract_excerpt":"Notable progress has been made in numerous fields of machine learning based on neural network-driven mutual information (MI) bounds. However, utilizing the conventional MI-based losses is often challenging due to their practical and mathematical limitations. In this work, we first identify the symptoms behind their instability: (1) the neural network not converging even after the loss seemed to converge, and (2) saturating neural network outputs causing the loss to diverge. We mitigate both issues by adding a novel regularization term to the existing losses. We theoretically and experimentally"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2011.07932","kind":"arxiv","version":4},"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/2011.07932/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":"2011.07932","created_at":"2026-07-05T04:32:48.284915+00:00"},{"alias_kind":"arxiv_version","alias_value":"2011.07932v4","created_at":"2026-07-05T04:32:48.284915+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2011.07932","created_at":"2026-07-05T04:32:48.284915+00:00"},{"alias_kind":"pith_short_12","alias_value":"VJTCAPDK56PU","created_at":"2026-07-05T04:32:48.284915+00:00"},{"alias_kind":"pith_short_16","alias_value":"VJTCAPDK56PU4XHC","created_at":"2026-07-05T04:32:48.284915+00:00"},{"alias_kind":"pith_short_8","alias_value":"VJTCAPDK","created_at":"2026-07-05T04:32:48.284915+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/VJTCAPDK56PU4XHCMSKUSIVQNL","json":"https://pith.science/pith/VJTCAPDK56PU4XHCMSKUSIVQNL.json","graph_json":"https://pith.science/api/pith-number/VJTCAPDK56PU4XHCMSKUSIVQNL/graph.json","events_json":"https://pith.science/api/pith-number/VJTCAPDK56PU4XHCMSKUSIVQNL/events.json","paper":"https://pith.science/paper/VJTCAPDK"},"agent_actions":{"view_html":"https://pith.science/pith/VJTCAPDK56PU4XHCMSKUSIVQNL","download_json":"https://pith.science/pith/VJTCAPDK56PU4XHCMSKUSIVQNL.json","view_paper":"https://pith.science/paper/VJTCAPDK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2011.07932&json=true","fetch_graph":"https://pith.science/api/pith-number/VJTCAPDK56PU4XHCMSKUSIVQNL/graph.json","fetch_events":"https://pith.science/api/pith-number/VJTCAPDK56PU4XHCMSKUSIVQNL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VJTCAPDK56PU4XHCMSKUSIVQNL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VJTCAPDK56PU4XHCMSKUSIVQNL/action/storage_attestation","attest_author":"https://pith.science/pith/VJTCAPDK56PU4XHCMSKUSIVQNL/action/author_attestation","sign_citation":"https://pith.science/pith/VJTCAPDK56PU4XHCMSKUSIVQNL/action/citation_signature","submit_replication":"https://pith.science/pith/VJTCAPDK56PU4XHCMSKUSIVQNL/action/replication_record"}},"created_at":"2026-07-05T04:32:48.284915+00:00","updated_at":"2026-07-05T04:32:48.284915+00:00"}