{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:36RQD25OJ62ZTK4M4FKYOEVVSN","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"d4ff8c789f0a7c9ba6e1a3594b721ea730b61cee8dc272f53cc5af9c63833d21","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-26T14:14:51Z","title_canon_sha256":"b7cb069c45549099864305f99713a0b3d740c92cefeb268d9915411db4efab42"},"schema_version":"1.0","source":{"id":"2305.16958","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.16958","created_at":"2026-07-05T08:23:09Z"},{"alias_kind":"arxiv_version","alias_value":"2305.16958v2","created_at":"2026-07-05T08:23:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.16958","created_at":"2026-07-05T08:23:09Z"},{"alias_kind":"pith_short_12","alias_value":"36RQD25OJ62Z","created_at":"2026-07-05T08:23:09Z"},{"alias_kind":"pith_short_16","alias_value":"36RQD25OJ62ZTK4M","created_at":"2026-07-05T08:23:09Z"},{"alias_kind":"pith_short_8","alias_value":"36RQD25O","created_at":"2026-07-05T08:23:09Z"}],"graph_snapshots":[{"event_id":"sha256:48a18f0d09b3e78aa41c88ce6dd10ccb9f2fb7e8f9d3733a771ea4b518d2f8e2","target":"graph","created_at":"2026-07-05T08:23:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2305.16958/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Autoregressive language models are trained by minimizing the cross-entropy of the model distribution Q relative to the data distribution P -- that is, minimizing the forward cross-entropy, which is equivalent to maximum likelihood estimation (MLE). We have observed that models trained in this way may \"over-generalize\", in the sense that they produce non-human-like text. Moreover, we believe that reverse cross-entropy, i.e., the cross-entropy of P relative to Q, is a better reflection of how a human would evaluate text generated by a model. Hence, we propose learning with MixCE, an objective th","authors_text":"David Rosenberg, Mark Dredze, Mohit Bansal, Ozan Irsoy, Shijie Wu, Shiyue Zhang, Steven Lu","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-26T14:14:51Z","title":"MixCE: Training Autoregressive Language Models by Mixing Forward and Reverse Cross-Entropies"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.16958","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:4fb6a89e6d0a7eecb8c96c020aa3ade3dd6c2bf93a35c2effa9abbcc6340b9c9","target":"record","created_at":"2026-07-05T08:23:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"d4ff8c789f0a7c9ba6e1a3594b721ea730b61cee8dc272f53cc5af9c63833d21","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-26T14:14:51Z","title_canon_sha256":"b7cb069c45549099864305f99713a0b3d740c92cefeb268d9915411db4efab42"},"schema_version":"1.0","source":{"id":"2305.16958","kind":"arxiv","version":2}},"canonical_sha256":"dfa301ebae4fb599ab8ce1558712b5935dfad0526ed8c241eaa77639729fc26b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"dfa301ebae4fb599ab8ce1558712b5935dfad0526ed8c241eaa77639729fc26b","first_computed_at":"2026-07-05T08:23:09.219689Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:23:09.219689Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"+zFL6NCf5TCSW1ATzflf+2eSxxReA/a8xc0iTsPLkTBWpCRtNb3VQHNyH6oHr6OtyldlLdZcpclTHAQDVasZDw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:23:09.220095Z","signed_message":"canonical_sha256_bytes"},"source_id":"2305.16958","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4fb6a89e6d0a7eecb8c96c020aa3ade3dd6c2bf93a35c2effa9abbcc6340b9c9","sha256:48a18f0d09b3e78aa41c88ce6dd10ccb9f2fb7e8f9d3733a771ea4b518d2f8e2"],"state_sha256":"0c69319861c4d1b4404d0e90bbd4098e2210d2ebc80365b3d6a3f65e0d7c40cf"}