{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:2IE2ALFJF4TDNRSCHRPG3IYJUU","short_pith_number":"pith:2IE2ALFJ","canonical_record":{"source":{"id":"2506.18203","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-22T23:38:15Z","cross_cats_sorted":[],"title_canon_sha256":"ac9bf8e947c75189bdc4d83ecd0bd8e072a6c357a434c098403414e44a6c670b","abstract_canon_sha256":"30b1972afb828116b1c7bf6246a283a79699603743c103b7e7aa7a030b920dec"},"schema_version":"1.0"},"canonical_sha256":"d209a02ca92f2636c6423c5e6da309a520dea4de66f9453405633389f0762559","source":{"kind":"arxiv","id":"2506.18203","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.18203","created_at":"2026-08-07T00:44:58Z"},{"alias_kind":"arxiv_version","alias_value":"2506.18203v3","created_at":"2026-08-07T00:44:58Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.18203","created_at":"2026-08-07T00:44:58Z"},{"alias_kind":"pith_short_12","alias_value":"2IE2ALFJF4TD","created_at":"2026-08-07T00:44:58Z"},{"alias_kind":"pith_short_16","alias_value":"2IE2ALFJF4TDNRSC","created_at":"2026-08-07T00:44:58Z"},{"alias_kind":"pith_short_8","alias_value":"2IE2ALFJ","created_at":"2026-08-07T00:44:58Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:2IE2ALFJF4TDNRSCHRPG3IYJUU","target":"record","payload":{"canonical_record":{"source":{"id":"2506.18203","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-22T23:38:15Z","cross_cats_sorted":[],"title_canon_sha256":"ac9bf8e947c75189bdc4d83ecd0bd8e072a6c357a434c098403414e44a6c670b","abstract_canon_sha256":"30b1972afb828116b1c7bf6246a283a79699603743c103b7e7aa7a030b920dec"},"schema_version":"1.0"},"canonical_sha256":"d209a02ca92f2636c6423c5e6da309a520dea4de66f9453405633389f0762559","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-07T00:44:58.955792Z","signature_b64":"qdx7nzJ4IRoVog+/Hw2VuCxAdpPkmIQP4pJsaWCToUPeBglB2v3H64L+vQAnNPm36oNLj58b8IUOKLyl2F8lBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d209a02ca92f2636c6423c5e6da309a520dea4de66f9453405633389f0762559","last_reissued_at":"2026-08-07T00:44:58.954271Z","signature_status":"signed_v1","first_computed_at":"2026-08-07T00:44:58.954271Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2506.18203","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-08-07T00:44:58Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eGD8Hm5u5Knw5WnzWsKxAXuZ0faBPW9+/bQw5YNOGRoTLGr2BFsVjNbW4Im9SKtPXkZaOX9hwngCy3XiioOzBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T12:20:48.805068Z"},"content_sha256":"f9cc9b11e4a9bd34b53758f063f62e1a8235d0d6e0b6d182bd03c5b77bf6e84c","schema_version":"1.0","event_id":"sha256:f9cc9b11e4a9bd34b53758f063f62e1a8235d0d6e0b6d182bd03c5b77bf6e84c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:2IE2ALFJF4TDNRSCHRPG3IYJUU","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Shrinking the Generation-Verification Gap with Weak Verifiers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Azalia Mirhoseini, Ben Athiwaratkun, Brendan McLaughlin, Christopher R\\'e, E. Kelly Buchanan, Frederic Sala, Jon Saad-Falcon, Mayee F. Chen, Scott Linderman, Shang Zhu, Tanvir Bhathal, Tzu-Heng Huang","submitted_at":"2025-06-22T23:38:15Z","abstract_excerpt":"Verifiers can improve language model capabilities by scoring and ranking responses from generated candidates. Currently, high-quality verifiers are either unscalable (e.g., humans) or limited in utility (e.g., tools like Lean). While LM judges and reward models have become broadly useful as general-purpose verifiers, a significant performance gap remains between them and oracle verifiers (verifiers with perfect accuracy). To help close this gap, we introduce Weaver, a framework for designing a strong verifier by combining multiple weak, imperfect verifiers. We find weighted ensembles of verifi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.18203","kind":"arxiv","version":3},"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/2506.18203/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-08-07T00:44:58Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"a8Qv2aCTsAB5lGFYM7avoHru3kLhx3cXD8BfqXRs+1NM7z5ClrT1H8azCYS0XakF/p+KGVuPASHFCwbDsvHSAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T12:20:48.805569Z"},"content_sha256":"44b8d33993b5f3a1ea4e4d74b01f0b4cf0bb81525b12696ba9bc3cc741427b3a","schema_version":"1.0","event_id":"sha256:44b8d33993b5f3a1ea4e4d74b01f0b4cf0bb81525b12696ba9bc3cc741427b3a"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/2IE2ALFJF4TDNRSCHRPG3IYJUU/bundle.json","state_url":"https://pith.science/pith/2IE2ALFJF4TDNRSCHRPG3IYJUU/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/2IE2ALFJF4TDNRSCHRPG3IYJUU/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-20T12:20:48Z","links":{"resolver":"https://pith.science/pith/2IE2ALFJF4TDNRSCHRPG3IYJUU","bundle":"https://pith.science/pith/2IE2ALFJF4TDNRSCHRPG3IYJUU/bundle.json","state":"https://pith.science/pith/2IE2ALFJF4TDNRSCHRPG3IYJUU/state.json","well_known_bundle":"https://pith.science/.well-known/pith/2IE2ALFJF4TDNRSCHRPG3IYJUU/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:2IE2ALFJF4TDNRSCHRPG3IYJUU","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":"30b1972afb828116b1c7bf6246a283a79699603743c103b7e7aa7a030b920dec","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-22T23:38:15Z","title_canon_sha256":"ac9bf8e947c75189bdc4d83ecd0bd8e072a6c357a434c098403414e44a6c670b"},"schema_version":"1.0","source":{"id":"2506.18203","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.18203","created_at":"2026-08-07T00:44:58Z"},{"alias_kind":"arxiv_version","alias_value":"2506.18203v3","created_at":"2026-08-07T00:44:58Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.18203","created_at":"2026-08-07T00:44:58Z"},{"alias_kind":"pith_short_12","alias_value":"2IE2ALFJF4TD","created_at":"2026-08-07T00:44:58Z"},{"alias_kind":"pith_short_16","alias_value":"2IE2ALFJF4TDNRSC","created_at":"2026-08-07T00:44:58Z"},{"alias_kind":"pith_short_8","alias_value":"2IE2ALFJ","created_at":"2026-08-07T00:44:58Z"}],"graph_snapshots":[{"event_id":"sha256:44b8d33993b5f3a1ea4e4d74b01f0b4cf0bb81525b12696ba9bc3cc741427b3a","target":"graph","created_at":"2026-08-07T00:44:58Z","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/2506.18203/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Verifiers can improve language model capabilities by scoring and ranking responses from generated candidates. Currently, high-quality verifiers are either unscalable (e.g., humans) or limited in utility (e.g., tools like Lean). While LM judges and reward models have become broadly useful as general-purpose verifiers, a significant performance gap remains between them and oracle verifiers (verifiers with perfect accuracy). To help close this gap, we introduce Weaver, a framework for designing a strong verifier by combining multiple weak, imperfect verifiers. We find weighted ensembles of verifi","authors_text":"Azalia Mirhoseini, Ben Athiwaratkun, Brendan McLaughlin, Christopher R\\'e, E. Kelly Buchanan, Frederic Sala, Jon Saad-Falcon, Mayee F. Chen, Scott Linderman, Shang Zhu, Tanvir Bhathal, Tzu-Heng Huang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-22T23:38:15Z","title":"Shrinking the Generation-Verification Gap with Weak Verifiers"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.18203","kind":"arxiv","version":3},"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:f9cc9b11e4a9bd34b53758f063f62e1a8235d0d6e0b6d182bd03c5b77bf6e84c","target":"record","created_at":"2026-08-07T00:44:58Z","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":"30b1972afb828116b1c7bf6246a283a79699603743c103b7e7aa7a030b920dec","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-22T23:38:15Z","title_canon_sha256":"ac9bf8e947c75189bdc4d83ecd0bd8e072a6c357a434c098403414e44a6c670b"},"schema_version":"1.0","source":{"id":"2506.18203","kind":"arxiv","version":3}},"canonical_sha256":"d209a02ca92f2636c6423c5e6da309a520dea4de66f9453405633389f0762559","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d209a02ca92f2636c6423c5e6da309a520dea4de66f9453405633389f0762559","first_computed_at":"2026-08-07T00:44:58.954271Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-07T00:44:58.954271Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"qdx7nzJ4IRoVog+/Hw2VuCxAdpPkmIQP4pJsaWCToUPeBglB2v3H64L+vQAnNPm36oNLj58b8IUOKLyl2F8lBg==","signature_status":"signed_v1","signed_at":"2026-08-07T00:44:58.955792Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.18203","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f9cc9b11e4a9bd34b53758f063f62e1a8235d0d6e0b6d182bd03c5b77bf6e84c","sha256:44b8d33993b5f3a1ea4e4d74b01f0b4cf0bb81525b12696ba9bc3cc741427b3a"],"state_sha256":"b568e07de51e263eab0e0ef8965ae54f8566c23e63da2e63956d4681775c3acd"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"nSmVPm9HRtvVHVlJWlWUsPYqYRy5Ob72WQgtRjRiT4CI/V5GdhwA6bECYQsa8oApuHqW6ce8EjJ3urrLvQS2Dw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-20T12:20:48.811255Z","bundle_sha256":"c13eb5d63d2d9b01d409b1bb421e5f33ea5d428f201a803a2730edf31e7eba93"}}