{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:NAPJYZLSGEDJI66IDF3JBM5TGG","short_pith_number":"pith:NAPJYZLS","canonical_record":{"source":{"id":"2501.13124","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-21T05:36:13Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"9e7eb7d16f3c591be335b2b62cf3e6cfbfe8b1cbdcdd0db500e1e6890d380b32","abstract_canon_sha256":"1297f30ef05010a0a8b00573fe19ee009faf8ee39790e6ad4524d7081ffce2b0"},"schema_version":"1.0"},"canonical_sha256":"681e9c65723106947bc8197690b3b331aca6dad48479e258ea0af268a665cd53","source":{"kind":"arxiv","id":"2501.13124","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.13124","created_at":"2026-07-05T10:04:23Z"},{"alias_kind":"arxiv_version","alias_value":"2501.13124v1","created_at":"2026-07-05T10:04:23Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.13124","created_at":"2026-07-05T10:04:23Z"},{"alias_kind":"pith_short_12","alias_value":"NAPJYZLSGEDJ","created_at":"2026-07-05T10:04:23Z"},{"alias_kind":"pith_short_16","alias_value":"NAPJYZLSGEDJI66I","created_at":"2026-07-05T10:04:23Z"},{"alias_kind":"pith_short_8","alias_value":"NAPJYZLS","created_at":"2026-07-05T10:04:23Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:NAPJYZLSGEDJI66IDF3JBM5TGG","target":"record","payload":{"canonical_record":{"source":{"id":"2501.13124","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-21T05:36:13Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"9e7eb7d16f3c591be335b2b62cf3e6cfbfe8b1cbdcdd0db500e1e6890d380b32","abstract_canon_sha256":"1297f30ef05010a0a8b00573fe19ee009faf8ee39790e6ad4524d7081ffce2b0"},"schema_version":"1.0"},"canonical_sha256":"681e9c65723106947bc8197690b3b331aca6dad48479e258ea0af268a665cd53","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:04:23.759135Z","signature_b64":"ULDhOr6L9Rp/pDXQpa/5IR+kEJkHeklvSUg3d8N+rj6mEsRQ4j/bb2iDw4ndydAHtp7C554alAd/MCIBLIWbCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"681e9c65723106947bc8197690b3b331aca6dad48479e258ea0af268a665cd53","last_reissued_at":"2026-07-05T10:04:23.758652Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:04:23.758652Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2501.13124","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":"2026-07-05T10:04:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uTAXZs6NllOFd94PPeAyad8HfaLlxehKKRZtOTyc0ZNDHqYX5jGJhF01uHPjp5bnHWRbeZQ8LJW/qi8A8atNCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T07:16:33.139541Z"},"content_sha256":"77ba10875313e9b8d78d81b8d4e65e9862b9370617c0d22052697ea5e6ad8f75","schema_version":"1.0","event_id":"sha256:77ba10875313e9b8d78d81b8d4e65e9862b9370617c0d22052697ea5e6ad8f75"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:NAPJYZLSGEDJI66IDF3JBM5TGG","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Debate Helps Weak-to-Strong Generalization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Fei Huang, Hao Lang, Yongbin Li","submitted_at":"2025-01-21T05:36:13Z","abstract_excerpt":"Common methods for aligning already-capable models with desired behavior rely on the ability of humans to provide supervision. However, future superhuman models will surpass the capability of humans. Therefore, humans will only be able to weakly supervise superhuman models. This expected deficiency of human evaluation would weaken the safety of future AI systems. Scalable oversight and weak-to-strong generalization are two complementary approaches to tackle this issue. In this paper, we attempt to combine the strengths of these two approaches to further improve alignment. Specifically, we inve"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.13124","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/2501.13124/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":"2026-07-05T10:04:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Tf8q074CzvBeS9MEErb+bgi4FMQwITwvxuS/HIF1uSQiv2k8Crlo15Gdq+plLRxgtSSecxsoaJTfFpvOkrayBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T07:16:33.140035Z"},"content_sha256":"c9f227bed342a243bf2570e2fb2b771cb4fc068a2d95c82f816b789dcc491f32","schema_version":"1.0","event_id":"sha256:c9f227bed342a243bf2570e2fb2b771cb4fc068a2d95c82f816b789dcc491f32"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/NAPJYZLSGEDJI66IDF3JBM5TGG/bundle.json","state_url":"https://pith.science/pith/NAPJYZLSGEDJI66IDF3JBM5TGG/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/NAPJYZLSGEDJI66IDF3JBM5TGG/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":"2026-08-07T07:16:33Z","links":{"resolver":"https://pith.science/pith/NAPJYZLSGEDJI66IDF3JBM5TGG","bundle":"https://pith.science/pith/NAPJYZLSGEDJI66IDF3JBM5TGG/bundle.json","state":"https://pith.science/pith/NAPJYZLSGEDJI66IDF3JBM5TGG/state.json","well_known_bundle":"https://pith.science/.well-known/pith/NAPJYZLSGEDJI66IDF3JBM5TGG/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:NAPJYZLSGEDJI66IDF3JBM5TGG","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":"1297f30ef05010a0a8b00573fe19ee009faf8ee39790e6ad4524d7081ffce2b0","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-21T05:36:13Z","title_canon_sha256":"9e7eb7d16f3c591be335b2b62cf3e6cfbfe8b1cbdcdd0db500e1e6890d380b32"},"schema_version":"1.0","source":{"id":"2501.13124","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.13124","created_at":"2026-07-05T10:04:23Z"},{"alias_kind":"arxiv_version","alias_value":"2501.13124v1","created_at":"2026-07-05T10:04:23Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.13124","created_at":"2026-07-05T10:04:23Z"},{"alias_kind":"pith_short_12","alias_value":"NAPJYZLSGEDJ","created_at":"2026-07-05T10:04:23Z"},{"alias_kind":"pith_short_16","alias_value":"NAPJYZLSGEDJI66I","created_at":"2026-07-05T10:04:23Z"},{"alias_kind":"pith_short_8","alias_value":"NAPJYZLS","created_at":"2026-07-05T10:04:23Z"}],"graph_snapshots":[{"event_id":"sha256:c9f227bed342a243bf2570e2fb2b771cb4fc068a2d95c82f816b789dcc491f32","target":"graph","created_at":"2026-07-05T10:04:23Z","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/2501.13124/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Common methods for aligning already-capable models with desired behavior rely on the ability of humans to provide supervision. However, future superhuman models will surpass the capability of humans. Therefore, humans will only be able to weakly supervise superhuman models. This expected deficiency of human evaluation would weaken the safety of future AI systems. Scalable oversight and weak-to-strong generalization are two complementary approaches to tackle this issue. In this paper, we attempt to combine the strengths of these two approaches to further improve alignment. Specifically, we inve","authors_text":"Fei Huang, Hao Lang, Yongbin Li","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-21T05:36:13Z","title":"Debate Helps Weak-to-Strong Generalization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.13124","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":"sha256:77ba10875313e9b8d78d81b8d4e65e9862b9370617c0d22052697ea5e6ad8f75","target":"record","created_at":"2026-07-05T10:04:23Z","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":"1297f30ef05010a0a8b00573fe19ee009faf8ee39790e6ad4524d7081ffce2b0","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-21T05:36:13Z","title_canon_sha256":"9e7eb7d16f3c591be335b2b62cf3e6cfbfe8b1cbdcdd0db500e1e6890d380b32"},"schema_version":"1.0","source":{"id":"2501.13124","kind":"arxiv","version":1}},"canonical_sha256":"681e9c65723106947bc8197690b3b331aca6dad48479e258ea0af268a665cd53","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"681e9c65723106947bc8197690b3b331aca6dad48479e258ea0af268a665cd53","first_computed_at":"2026-07-05T10:04:23.758652Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:04:23.758652Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ULDhOr6L9Rp/pDXQpa/5IR+kEJkHeklvSUg3d8N+rj6mEsRQ4j/bb2iDw4ndydAHtp7C554alAd/MCIBLIWbCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:04:23.759135Z","signed_message":"canonical_sha256_bytes"},"source_id":"2501.13124","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:77ba10875313e9b8d78d81b8d4e65e9862b9370617c0d22052697ea5e6ad8f75","sha256:c9f227bed342a243bf2570e2fb2b771cb4fc068a2d95c82f816b789dcc491f32"],"state_sha256":"e664f407a853c3e04431401b343c2621887d5baa067a533207a3f0e1a020482d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eE6AUrU+TMHgHz+M6kuZhUpIeqrpT8hrrLUQzRUI/enwkRUPROW+oDfp7yFvJ0Ch6YnT0m4CBZCjWc9AYR7BBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T07:16:33.145301Z","bundle_sha256":"13fe1f00a9d74ca1f86cdc06247b70bd0a576694e7203a033e35b87cb568029c"}}