{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:SAD3VLBWBP2QX2Z6F5O3JEXDEI","short_pith_number":"pith:SAD3VLBW","canonical_record":{"source":{"id":"2605.25629","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-25T09:30:49Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"7303dda99b4f2611f3bbc2a7ffe49fac1d14734dbf1b716bdf61961d9af878f0","abstract_canon_sha256":"aee6572a9d7cffc2dd11ba39cfb9ee20a9168bb047b7cb6e14bf266fb6af11ae"},"schema_version":"1.0"},"canonical_sha256":"9007baac360bf50beb3e2f5db492e322302febdf125e5db9c3c44a55b46329b2","source":{"kind":"arxiv","id":"2605.25629","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.25629","created_at":"2026-05-26T02:04:47Z"},{"alias_kind":"arxiv_version","alias_value":"2605.25629v1","created_at":"2026-05-26T02:04:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.25629","created_at":"2026-05-26T02:04:47Z"},{"alias_kind":"pith_short_12","alias_value":"SAD3VLBWBP2Q","created_at":"2026-05-26T02:04:47Z"},{"alias_kind":"pith_short_16","alias_value":"SAD3VLBWBP2QX2Z6","created_at":"2026-05-26T02:04:47Z"},{"alias_kind":"pith_short_8","alias_value":"SAD3VLBW","created_at":"2026-05-26T02:04:47Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:SAD3VLBWBP2QX2Z6F5O3JEXDEI","target":"record","payload":{"canonical_record":{"source":{"id":"2605.25629","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-25T09:30:49Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"7303dda99b4f2611f3bbc2a7ffe49fac1d14734dbf1b716bdf61961d9af878f0","abstract_canon_sha256":"aee6572a9d7cffc2dd11ba39cfb9ee20a9168bb047b7cb6e14bf266fb6af11ae"},"schema_version":"1.0"},"canonical_sha256":"9007baac360bf50beb3e2f5db492e322302febdf125e5db9c3c44a55b46329b2","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T02:04:47.276420Z","signature_b64":"V0YB4jUx44l5xFT01NA+um1MPgfxuy45Dg37evCWZg/i9pBa3vtduEGpxWyhSCXUyCdBKQWqNHsI8wYUrVObDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9007baac360bf50beb3e2f5db492e322302febdf125e5db9c3c44a55b46329b2","last_reissued_at":"2026-05-26T02:04:47.275821Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T02:04:47.275821Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.25629","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":"2026-05-26T02:04:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZRYmj6/iIKP1QiwzFSUKd1pxvBLZSps32UArToRPcfVb05qxqj3IG12Wf6ssQj8uV+f3Qla10fTrkJxPeovRBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T15:23:50.859815Z"},"content_sha256":"a35965bb5b68303466b57bc8a97705396b9d487b49c8502ebc143b46286cc3b9","schema_version":"1.0","event_id":"sha256:a35965bb5b68303466b57bc8a97705396b9d487b49c8502ebc143b46286cc3b9"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:SAD3VLBWBP2QX2Z6F5O3JEXDEI","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"When In-Distribution Gains Fail: Evaluating Weak-to-Strong Reward Models under Preference Shift","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Anh Tuan Luu, Cong-Duy Nguyen, Khoi Le, Miao Chunyan, Phong Nguyen, See-kiong Ng, Thong Nguyen, Tri Cao","submitted_at":"2026-05-25T09:30:49Z","abstract_excerpt":"Weak-to-strong (W2S) generalization is a promising framework for scalable oversight, yet existing evaluations often test students under matched train--test distributions. Therefore, we study W2S preference learning under zero-shot distribution shift and find that strong students trained on weak preference labels can appear successful in-distribution while failing to transfer across preference datasets. We provide evidence for a representational failure mode in which weak-supervised fine-tuning can pull the strong model toward source-domain features instead of maintaining broadly transferable p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.25629","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/2605.25629/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":"2026-05-26T02:04:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"41xRA5E+1vf2JQ16IDdbA+YMKQHUbKg3bbzuihP/ginVF6rjQ2StCmaeLH897iXhm9jzWK8xBtjE4zzXz1f6Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T15:23:50.860592Z"},"content_sha256":"ea0073ca3bcd24b99e694a97dffd8376aa8ad62a8edb62bf2c9e97ff64d3de89","schema_version":"1.0","event_id":"sha256:ea0073ca3bcd24b99e694a97dffd8376aa8ad62a8edb62bf2c9e97ff64d3de89"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/SAD3VLBWBP2QX2Z6F5O3JEXDEI/bundle.json","state_url":"https://pith.science/pith/SAD3VLBWBP2QX2Z6F5O3JEXDEI/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/SAD3VLBWBP2QX2Z6F5O3JEXDEI/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":"2026-05-26T15:23:50Z","links":{"resolver":"https://pith.science/pith/SAD3VLBWBP2QX2Z6F5O3JEXDEI","bundle":"https://pith.science/pith/SAD3VLBWBP2QX2Z6F5O3JEXDEI/bundle.json","state":"https://pith.science/pith/SAD3VLBWBP2QX2Z6F5O3JEXDEI/state.json","well_known_bundle":"https://pith.science/.well-known/pith/SAD3VLBWBP2QX2Z6F5O3JEXDEI/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:SAD3VLBWBP2QX2Z6F5O3JEXDEI","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":"aee6572a9d7cffc2dd11ba39cfb9ee20a9168bb047b7cb6e14bf266fb6af11ae","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-25T09:30:49Z","title_canon_sha256":"7303dda99b4f2611f3bbc2a7ffe49fac1d14734dbf1b716bdf61961d9af878f0"},"schema_version":"1.0","source":{"id":"2605.25629","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.25629","created_at":"2026-05-26T02:04:47Z"},{"alias_kind":"arxiv_version","alias_value":"2605.25629v1","created_at":"2026-05-26T02:04:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.25629","created_at":"2026-05-26T02:04:47Z"},{"alias_kind":"pith_short_12","alias_value":"SAD3VLBWBP2Q","created_at":"2026-05-26T02:04:47Z"},{"alias_kind":"pith_short_16","alias_value":"SAD3VLBWBP2QX2Z6","created_at":"2026-05-26T02:04:47Z"},{"alias_kind":"pith_short_8","alias_value":"SAD3VLBW","created_at":"2026-05-26T02:04:47Z"}],"graph_snapshots":[{"event_id":"sha256:ea0073ca3bcd24b99e694a97dffd8376aa8ad62a8edb62bf2c9e97ff64d3de89","target":"graph","created_at":"2026-05-26T02:04:47Z","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/2605.25629/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Weak-to-strong (W2S) generalization is a promising framework for scalable oversight, yet existing evaluations often test students under matched train--test distributions. Therefore, we study W2S preference learning under zero-shot distribution shift and find that strong students trained on weak preference labels can appear successful in-distribution while failing to transfer across preference datasets. We provide evidence for a representational failure mode in which weak-supervised fine-tuning can pull the strong model toward source-domain features instead of maintaining broadly transferable p","authors_text":"Anh Tuan Luu, Cong-Duy Nguyen, Khoi Le, Miao Chunyan, Phong Nguyen, See-kiong Ng, Thong Nguyen, Tri Cao","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-25T09:30:49Z","title":"When In-Distribution Gains Fail: Evaluating Weak-to-Strong Reward Models under Preference Shift"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.25629","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:a35965bb5b68303466b57bc8a97705396b9d487b49c8502ebc143b46286cc3b9","target":"record","created_at":"2026-05-26T02:04:47Z","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":"aee6572a9d7cffc2dd11ba39cfb9ee20a9168bb047b7cb6e14bf266fb6af11ae","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-25T09:30:49Z","title_canon_sha256":"7303dda99b4f2611f3bbc2a7ffe49fac1d14734dbf1b716bdf61961d9af878f0"},"schema_version":"1.0","source":{"id":"2605.25629","kind":"arxiv","version":1}},"canonical_sha256":"9007baac360bf50beb3e2f5db492e322302febdf125e5db9c3c44a55b46329b2","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9007baac360bf50beb3e2f5db492e322302febdf125e5db9c3c44a55b46329b2","first_computed_at":"2026-05-26T02:04:47.275821Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-26T02:04:47.275821Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"V0YB4jUx44l5xFT01NA+um1MPgfxuy45Dg37evCWZg/i9pBa3vtduEGpxWyhSCXUyCdBKQWqNHsI8wYUrVObDQ==","signature_status":"signed_v1","signed_at":"2026-05-26T02:04:47.276420Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.25629","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a35965bb5b68303466b57bc8a97705396b9d487b49c8502ebc143b46286cc3b9","sha256:ea0073ca3bcd24b99e694a97dffd8376aa8ad62a8edb62bf2c9e97ff64d3de89"],"state_sha256":"b49269ce47f9f798df849675e86c14ab1e6f3d5eee754435803741cf7129ef72"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"b1qeMu+4tKSYYXL1xLW5bQjOeQIdIwrTUpaHb7G+81PQnSYDc8ZbSurmFFzJem4LkQkGFU63PERr0hfkaqMfDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-26T15:23:50.864115Z","bundle_sha256":"bdc57fc93227ce1f22491c03976af43713ce9030bd00b7c94f2c5c51d696a862"}}