{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:PCXIQQCA2M7JL4S6N5WKWB5QCF","short_pith_number":"pith:PCXIQQCA","schema_version":"1.0","canonical_sha256":"78ae884040d33e95f25e6f6cab07b011492ee66ab609c24b7a92ecdbdb6f0891","source":{"kind":"arxiv","id":"2506.04609","version":1},"attestation_state":"computed","paper":{"title":"Exploring bidirectional bounds for minimax-training of Energy-based models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Cong Geng, Jes Frellsen, Jia Wang, Li Chen, S{\\o}ren Hauberg, Zhiyong Gao","submitted_at":"2025-06-05T03:58:03Z","abstract_excerpt":"Energy-based models (EBMs) estimate unnormalized densities in an elegant framework, but they are generally difficult to train. Recent work has linked EBMs to generative adversarial networks, by noting that they can be trained through a minimax game using a variational lower bound. To avoid the instabilities caused by minimizing a lower bound, we propose to instead work with bidirectional bounds, meaning that we maximize a lower bound and minimize an upper bound when training the EBM. We investigate four different bounds on the log-likelihood derived from different perspectives. We derive lower"},"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":"2506.04609","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-06-05T03:58:03Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"e273b2e7af878fc25f4a2c8bfdec7c4c32636c7acb38a0b40a7b0317c4fbe8bb","abstract_canon_sha256":"79edbf260b671b3c7aa3fcd0be24c04610942dec10ba5c495a8237c6a94564e0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:16:17.777575Z","signature_b64":"T5dqizRcXtVEh1q3frwl1CIloR06gd4M1fzXwmn5zE55CXbaCkmGoXv7cU2FOPHf7Uo7dfK3INfkqB4QxxZCBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"78ae884040d33e95f25e6f6cab07b011492ee66ab609c24b7a92ecdbdb6f0891","last_reissued_at":"2026-07-05T11:16:17.777043Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:16:17.777043Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Exploring bidirectional bounds for minimax-training of Energy-based models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Cong Geng, Jes Frellsen, Jia Wang, Li Chen, S{\\o}ren Hauberg, Zhiyong Gao","submitted_at":"2025-06-05T03:58:03Z","abstract_excerpt":"Energy-based models (EBMs) estimate unnormalized densities in an elegant framework, but they are generally difficult to train. Recent work has linked EBMs to generative adversarial networks, by noting that they can be trained through a minimax game using a variational lower bound. To avoid the instabilities caused by minimizing a lower bound, we propose to instead work with bidirectional bounds, meaning that we maximize a lower bound and minimize an upper bound when training the EBM. We investigate four different bounds on the log-likelihood derived from different perspectives. We derive lower"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.04609","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/2506.04609/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":"2506.04609","created_at":"2026-07-05T11:16:17.777098+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.04609v1","created_at":"2026-07-05T11:16:17.777098+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.04609","created_at":"2026-07-05T11:16:17.777098+00:00"},{"alias_kind":"pith_short_12","alias_value":"PCXIQQCA2M7J","created_at":"2026-07-05T11:16:17.777098+00:00"},{"alias_kind":"pith_short_16","alias_value":"PCXIQQCA2M7JL4S6","created_at":"2026-07-05T11:16:17.777098+00:00"},{"alias_kind":"pith_short_8","alias_value":"PCXIQQCA","created_at":"2026-07-05T11:16:17.777098+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/PCXIQQCA2M7JL4S6N5WKWB5QCF","json":"https://pith.science/pith/PCXIQQCA2M7JL4S6N5WKWB5QCF.json","graph_json":"https://pith.science/api/pith-number/PCXIQQCA2M7JL4S6N5WKWB5QCF/graph.json","events_json":"https://pith.science/api/pith-number/PCXIQQCA2M7JL4S6N5WKWB5QCF/events.json","paper":"https://pith.science/paper/PCXIQQCA"},"agent_actions":{"view_html":"https://pith.science/pith/PCXIQQCA2M7JL4S6N5WKWB5QCF","download_json":"https://pith.science/pith/PCXIQQCA2M7JL4S6N5WKWB5QCF.json","view_paper":"https://pith.science/paper/PCXIQQCA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.04609&json=true","fetch_graph":"https://pith.science/api/pith-number/PCXIQQCA2M7JL4S6N5WKWB5QCF/graph.json","fetch_events":"https://pith.science/api/pith-number/PCXIQQCA2M7JL4S6N5WKWB5QCF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PCXIQQCA2M7JL4S6N5WKWB5QCF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PCXIQQCA2M7JL4S6N5WKWB5QCF/action/storage_attestation","attest_author":"https://pith.science/pith/PCXIQQCA2M7JL4S6N5WKWB5QCF/action/author_attestation","sign_citation":"https://pith.science/pith/PCXIQQCA2M7JL4S6N5WKWB5QCF/action/citation_signature","submit_replication":"https://pith.science/pith/PCXIQQCA2M7JL4S6N5WKWB5QCF/action/replication_record"}},"created_at":"2026-07-05T11:16:17.777098+00:00","updated_at":"2026-07-05T11:16:17.777098+00:00"}