{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:ULSATT5RUCQCT452QMGZ6PZQXT","short_pith_number":"pith:ULSATT5R","canonical_record":{"source":{"id":"2606.27578","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-25T22:09:50Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"b15f1f6e47d5c21dbc284a270a76659efbfd72787380f5960b78bfc1bace69a1","abstract_canon_sha256":"7a7df38a870156ad05e8f7f9be830f8ba7a663ea62dee40950b759861a152c25"},"schema_version":"1.0"},"canonical_sha256":"a2e409cfb1a0a029f3ba830d9f3f30bcc535a3c12ac59a86708fe91b2c202d55","source":{"kind":"arxiv","id":"2606.27578","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.27578","created_at":"2026-06-29T00:14:10Z"},{"alias_kind":"arxiv_version","alias_value":"2606.27578v1","created_at":"2026-06-29T00:14:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.27578","created_at":"2026-06-29T00:14:10Z"},{"alias_kind":"pith_short_12","alias_value":"ULSATT5RUCQC","created_at":"2026-06-29T00:14:10Z"},{"alias_kind":"pith_short_16","alias_value":"ULSATT5RUCQCT452","created_at":"2026-06-29T00:14:10Z"},{"alias_kind":"pith_short_8","alias_value":"ULSATT5R","created_at":"2026-06-29T00:14:10Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:ULSATT5RUCQCT452QMGZ6PZQXT","target":"record","payload":{"canonical_record":{"source":{"id":"2606.27578","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-25T22:09:50Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"b15f1f6e47d5c21dbc284a270a76659efbfd72787380f5960b78bfc1bace69a1","abstract_canon_sha256":"7a7df38a870156ad05e8f7f9be830f8ba7a663ea62dee40950b759861a152c25"},"schema_version":"1.0"},"canonical_sha256":"a2e409cfb1a0a029f3ba830d9f3f30bcc535a3c12ac59a86708fe91b2c202d55","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-29T00:14:10.538721Z","signature_b64":"fpn2ntGbcygXPDQu6NJru+KJ0CFcrZiVrPZA6HQKIKxI4a+s3xVuGjALWA0YepVJ2nND6KAxM99SjLuld8/KAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a2e409cfb1a0a029f3ba830d9f3f30bcc535a3c12ac59a86708fe91b2c202d55","last_reissued_at":"2026-06-29T00:14:10.538358Z","signature_status":"signed_v1","first_computed_at":"2026-06-29T00:14:10.538358Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.27578","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-29T00:14:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hkUDy0SsnB72/3YkYo0hSpm4nvNP+zmqSqpfO+Esa3kzivRPLaDz6r5aYPm/4tQqe4ARXNdrx+7xZ35+eCh9DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T19:30:31.580970Z"},"content_sha256":"0f9f5dc558b23b26d5360ed883a586058e3d2f8cb6b0396c62385502467683f3","schema_version":"1.0","event_id":"sha256:0f9f5dc558b23b26d5360ed883a586058e3d2f8cb6b0396c62385502467683f3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:ULSATT5RUCQCT452QMGZ6PZQXT","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"PEBS: Per-rater Empirical-Bayes Shrinkage for RLHF Reward-Model Calibration","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Arnav Raj","submitted_at":"2026-06-25T22:09:50Z","abstract_excerpt":"Reward models for Reinforcement Learning from Human Feedback (RLHF) pool preferences across thousands of annotators and fit one global affine calibrator, collapsing raters with systematically different rating-scale offsets and slopes into a single average-rater fit that does not match any individual annotator. PEBS is a per-rater empirical-Bayes shrinkage estimator: it fits per-rater affine calibrators on a held-out slice of each annotator's ratings and applies Morris-James-Stein empirical-Bayes shrinkage toward the population mean, in closed form and without retraining the reward model. On PR"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.27578","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/2606.27578/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":"2026-06-29T00:14:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zDbLswdoS7kND01g+paSUorcfQvtHGL19BnVKcoPC+n2GbS9fZ8quQEVp5QLRmz38V4zxOVmMJPX694j4A4SAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T19:30:31.581396Z"},"content_sha256":"46ec3f6f98b59d6d306e0aa9a4a1ed32076ce0b9837204c9f522b43cafabd129","schema_version":"1.0","event_id":"sha256:46ec3f6f98b59d6d306e0aa9a4a1ed32076ce0b9837204c9f522b43cafabd129"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ULSATT5RUCQCT452QMGZ6PZQXT/bundle.json","state_url":"https://pith.science/pith/ULSATT5RUCQCT452QMGZ6PZQXT/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ULSATT5RUCQCT452QMGZ6PZQXT/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":"2026-06-29T19:30:31Z","links":{"resolver":"https://pith.science/pith/ULSATT5RUCQCT452QMGZ6PZQXT","bundle":"https://pith.science/pith/ULSATT5RUCQCT452QMGZ6PZQXT/bundle.json","state":"https://pith.science/pith/ULSATT5RUCQCT452QMGZ6PZQXT/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ULSATT5RUCQCT452QMGZ6PZQXT/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:ULSATT5RUCQCT452QMGZ6PZQXT","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":"7a7df38a870156ad05e8f7f9be830f8ba7a663ea62dee40950b759861a152c25","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-25T22:09:50Z","title_canon_sha256":"b15f1f6e47d5c21dbc284a270a76659efbfd72787380f5960b78bfc1bace69a1"},"schema_version":"1.0","source":{"id":"2606.27578","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.27578","created_at":"2026-06-29T00:14:10Z"},{"alias_kind":"arxiv_version","alias_value":"2606.27578v1","created_at":"2026-06-29T00:14:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.27578","created_at":"2026-06-29T00:14:10Z"},{"alias_kind":"pith_short_12","alias_value":"ULSATT5RUCQC","created_at":"2026-06-29T00:14:10Z"},{"alias_kind":"pith_short_16","alias_value":"ULSATT5RUCQCT452","created_at":"2026-06-29T00:14:10Z"},{"alias_kind":"pith_short_8","alias_value":"ULSATT5R","created_at":"2026-06-29T00:14:10Z"}],"graph_snapshots":[{"event_id":"sha256:46ec3f6f98b59d6d306e0aa9a4a1ed32076ce0b9837204c9f522b43cafabd129","target":"graph","created_at":"2026-06-29T00:14:10Z","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/2606.27578/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reward models for Reinforcement Learning from Human Feedback (RLHF) pool preferences across thousands of annotators and fit one global affine calibrator, collapsing raters with systematically different rating-scale offsets and slopes into a single average-rater fit that does not match any individual annotator. PEBS is a per-rater empirical-Bayes shrinkage estimator: it fits per-rater affine calibrators on a held-out slice of each annotator's ratings and applies Morris-James-Stein empirical-Bayes shrinkage toward the population mean, in closed form and without retraining the reward model. On PR","authors_text":"Arnav Raj","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-25T22:09:50Z","title":"PEBS: Per-rater Empirical-Bayes Shrinkage for RLHF Reward-Model Calibration"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.27578","kind":"arxiv","version":1},"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":"sha256:0f9f5dc558b23b26d5360ed883a586058e3d2f8cb6b0396c62385502467683f3","target":"record","created_at":"2026-06-29T00:14:10Z","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":"7a7df38a870156ad05e8f7f9be830f8ba7a663ea62dee40950b759861a152c25","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-25T22:09:50Z","title_canon_sha256":"b15f1f6e47d5c21dbc284a270a76659efbfd72787380f5960b78bfc1bace69a1"},"schema_version":"1.0","source":{"id":"2606.27578","kind":"arxiv","version":1}},"canonical_sha256":"a2e409cfb1a0a029f3ba830d9f3f30bcc535a3c12ac59a86708fe91b2c202d55","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a2e409cfb1a0a029f3ba830d9f3f30bcc535a3c12ac59a86708fe91b2c202d55","first_computed_at":"2026-06-29T00:14:10.538358Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-29T00:14:10.538358Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"fpn2ntGbcygXPDQu6NJru+KJ0CFcrZiVrPZA6HQKIKxI4a+s3xVuGjALWA0YepVJ2nND6KAxM99SjLuld8/KAg==","signature_status":"signed_v1","signed_at":"2026-06-29T00:14:10.538721Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.27578","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0f9f5dc558b23b26d5360ed883a586058e3d2f8cb6b0396c62385502467683f3","sha256:46ec3f6f98b59d6d306e0aa9a4a1ed32076ce0b9837204c9f522b43cafabd129"],"state_sha256":"cfdd1067afa9bb481cc8ada6f96fd1e5660c581ba1dd4b29eaaf3b57f438f12d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9VysVbYAQXIlia72uqg1/6WUNQFbe9Y5AyMqMBaCbtoW8chOQaWmBziKnKQgLo+yphnN7G0YiXmB2O2VUySPDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-29T19:30:31.583367Z","bundle_sha256":"7c4c29dce0dcc90f13765eb859b6d79ae6c9bc38df6356111a27423910e2411f"}}