{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:2BCEDVT3WOXEMTBGIUJISANF5O","short_pith_number":"pith:2BCEDVT3","schema_version":"1.0","canonical_sha256":"d04441d67bb3ae464c2645128901a5ebb2765e284491033927d1f50515748048","source":{"kind":"arxiv","id":"2607.23333","version":1},"attestation_state":"computed","paper":{"title":"Training with (Swap) Regret Loss in a Single-Layer Self-Attention Model: A Case Study on the Probability Simplex","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Asuman Ozdaglar, Chanwoo Park","submitted_at":"2026-07-25T19:11:15Z","abstract_excerpt":"We revisit the regret loss framework introduced in Park et al. (2025), which uses decision-theoretic regret as a direct loss function for training models to make better decisions, through the lens of probability-simplex policies. Our first result shows that a single-layer self-attention model trained with regret loss admits a stationary point whose forward-pass exactly matches smoothed fictitious play with the appropriate stepsize that ensures no-regret behavior-i.e., for any given policy input, the model outputs the same update that smoothed fictitious play would produce. In parallel, we also"},"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":"2607.23333","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-25T19:11:15Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"09d90817ed083202552767db5bc27dc8d2e4a490db473d33796de30fd282e182","abstract_canon_sha256":"068d36beb51bccf7c74a4604f3ca05960c3dc87586c363eed29ad7ea31c1d221"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-28T01:22:48.133467Z","signature_b64":"plG0SEmkgQxb1DQrGmYV1CfAJlYtRJeAsFvP51aaitMkdwH45W/vJFUDq6Y844zogvhl7C7VH1wmrmo4iRwZAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d04441d67bb3ae464c2645128901a5ebb2765e284491033927d1f50515748048","last_reissued_at":"2026-07-28T01:22:48.132727Z","signature_status":"signed_v1","first_computed_at":"2026-07-28T01:22:48.132727Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Training with (Swap) Regret Loss in a Single-Layer Self-Attention Model: A Case Study on the Probability Simplex","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Asuman Ozdaglar, Chanwoo Park","submitted_at":"2026-07-25T19:11:15Z","abstract_excerpt":"We revisit the regret loss framework introduced in Park et al. (2025), which uses decision-theoretic regret as a direct loss function for training models to make better decisions, through the lens of probability-simplex policies. Our first result shows that a single-layer self-attention model trained with regret loss admits a stationary point whose forward-pass exactly matches smoothed fictitious play with the appropriate stepsize that ensures no-regret behavior-i.e., for any given policy input, the model outputs the same update that smoothed fictitious play would produce. In parallel, we also"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.23333","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/2607.23333/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":"2607.23333","created_at":"2026-07-28T01:22:48.133118+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.23333v1","created_at":"2026-07-28T01:22:48.133118+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.23333","created_at":"2026-07-28T01:22:48.133118+00:00"},{"alias_kind":"pith_short_12","alias_value":"2BCEDVT3WOXE","created_at":"2026-07-28T01:22:48.133118+00:00"},{"alias_kind":"pith_short_16","alias_value":"2BCEDVT3WOXEMTBG","created_at":"2026-07-28T01:22:48.133118+00:00"},{"alias_kind":"pith_short_8","alias_value":"2BCEDVT3","created_at":"2026-07-28T01:22:48.133118+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/2BCEDVT3WOXEMTBGIUJISANF5O","json":"https://pith.science/pith/2BCEDVT3WOXEMTBGIUJISANF5O.json","graph_json":"https://pith.science/api/pith-number/2BCEDVT3WOXEMTBGIUJISANF5O/graph.json","events_json":"https://pith.science/api/pith-number/2BCEDVT3WOXEMTBGIUJISANF5O/events.json","paper":"https://pith.science/paper/2BCEDVT3"},"agent_actions":{"view_html":"https://pith.science/pith/2BCEDVT3WOXEMTBGIUJISANF5O","download_json":"https://pith.science/pith/2BCEDVT3WOXEMTBGIUJISANF5O.json","view_paper":"https://pith.science/paper/2BCEDVT3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.23333&json=true","fetch_graph":"https://pith.science/api/pith-number/2BCEDVT3WOXEMTBGIUJISANF5O/graph.json","fetch_events":"https://pith.science/api/pith-number/2BCEDVT3WOXEMTBGIUJISANF5O/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2BCEDVT3WOXEMTBGIUJISANF5O/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2BCEDVT3WOXEMTBGIUJISANF5O/action/storage_attestation","attest_author":"https://pith.science/pith/2BCEDVT3WOXEMTBGIUJISANF5O/action/author_attestation","sign_citation":"https://pith.science/pith/2BCEDVT3WOXEMTBGIUJISANF5O/action/citation_signature","submit_replication":"https://pith.science/pith/2BCEDVT3WOXEMTBGIUJISANF5O/action/replication_record"}},"created_at":"2026-07-28T01:22:48.133118+00:00","updated_at":"2026-07-28T01:22:48.133118+00:00"}