{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:3M3KGHQEOIHDQI35CMQGVPR3PZ","short_pith_number":"pith:3M3KGHQE","schema_version":"1.0","canonical_sha256":"db36a31e04720e38237d13206abe3b7e444532cf87315aa4a29ef52a1d956784","source":{"kind":"arxiv","id":"2606.31630","version":1},"attestation_state":"computed","paper":{"title":"Calibration, Not Compilation: Detecting and Repairing Misspecified Probabilistic Programs Written by Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Delu Zeng, Jian Xu, John Paisley, Qibin Zhao","submitted_at":"2026-06-30T13:16:39Z","abstract_excerpt":"Language models increasingly write probabilistic programs (in NumPyro, Stan, or Pyro), but a program that compiles, runs, and passes every unit test can still be \\emph{statistically} wrong -- a Gaussian likelihood for heavy-tailed data, a Poisson for over-dispersed counts, an invalid prior support, or a pathological parameterization. The right verifier is therefore not a test suite but the Bayesian workflow itself: posterior predictive checks, simulation-based calibration, sampler diagnostics ($\\hat R$, divergences, ESS), and held-out predictive density. We study this calibration oracle along "},"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":"2606.31630","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-30T13:16:39Z","cross_cats_sorted":[],"title_canon_sha256":"0e40e0dcc13125fcc5d71cb0ca3de3f9b112d8651bd2de373ce4fa1a247372f4","abstract_canon_sha256":"cdc925f026d2b2f0c29cba6deaec5c0105eb700a3a51abb6fd6a5a4008b43817"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-01T01:18:09.776855Z","signature_b64":"RWqtVhN4UsxFHH9rTyMWIUfi3uJe5Zuv07rXoWZoYaBC9NrwyNSt/uP+n76J/hAB9r9TG1mMhU1AfkcSp9pOBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"db36a31e04720e38237d13206abe3b7e444532cf87315aa4a29ef52a1d956784","last_reissued_at":"2026-07-01T01:18:09.776387Z","signature_status":"signed_v1","first_computed_at":"2026-07-01T01:18:09.776387Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Calibration, Not Compilation: Detecting and Repairing Misspecified Probabilistic Programs Written by Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Delu Zeng, Jian Xu, John Paisley, Qibin Zhao","submitted_at":"2026-06-30T13:16:39Z","abstract_excerpt":"Language models increasingly write probabilistic programs (in NumPyro, Stan, or Pyro), but a program that compiles, runs, and passes every unit test can still be \\emph{statistically} wrong -- a Gaussian likelihood for heavy-tailed data, a Poisson for over-dispersed counts, an invalid prior support, or a pathological parameterization. The right verifier is therefore not a test suite but the Bayesian workflow itself: posterior predictive checks, simulation-based calibration, sampler diagnostics ($\\hat R$, divergences, ESS), and held-out predictive density. We study this calibration oracle along "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.31630","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/2606.31630/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":"2606.31630","created_at":"2026-07-01T01:18:09.776442+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.31630v1","created_at":"2026-07-01T01:18:09.776442+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.31630","created_at":"2026-07-01T01:18:09.776442+00:00"},{"alias_kind":"pith_short_12","alias_value":"3M3KGHQEOIHD","created_at":"2026-07-01T01:18:09.776442+00:00"},{"alias_kind":"pith_short_16","alias_value":"3M3KGHQEOIHDQI35","created_at":"2026-07-01T01:18:09.776442+00:00"},{"alias_kind":"pith_short_8","alias_value":"3M3KGHQE","created_at":"2026-07-01T01:18:09.776442+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/3M3KGHQEOIHDQI35CMQGVPR3PZ","json":"https://pith.science/pith/3M3KGHQEOIHDQI35CMQGVPR3PZ.json","graph_json":"https://pith.science/api/pith-number/3M3KGHQEOIHDQI35CMQGVPR3PZ/graph.json","events_json":"https://pith.science/api/pith-number/3M3KGHQEOIHDQI35CMQGVPR3PZ/events.json","paper":"https://pith.science/paper/3M3KGHQE"},"agent_actions":{"view_html":"https://pith.science/pith/3M3KGHQEOIHDQI35CMQGVPR3PZ","download_json":"https://pith.science/pith/3M3KGHQEOIHDQI35CMQGVPR3PZ.json","view_paper":"https://pith.science/paper/3M3KGHQE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.31630&json=true","fetch_graph":"https://pith.science/api/pith-number/3M3KGHQEOIHDQI35CMQGVPR3PZ/graph.json","fetch_events":"https://pith.science/api/pith-number/3M3KGHQEOIHDQI35CMQGVPR3PZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3M3KGHQEOIHDQI35CMQGVPR3PZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3M3KGHQEOIHDQI35CMQGVPR3PZ/action/storage_attestation","attest_author":"https://pith.science/pith/3M3KGHQEOIHDQI35CMQGVPR3PZ/action/author_attestation","sign_citation":"https://pith.science/pith/3M3KGHQEOIHDQI35CMQGVPR3PZ/action/citation_signature","submit_replication":"https://pith.science/pith/3M3KGHQEOIHDQI35CMQGVPR3PZ/action/replication_record"}},"created_at":"2026-07-01T01:18:09.776442+00:00","updated_at":"2026-07-01T01:18:09.776442+00:00"}