{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:M3NGTFXYNYUZ2ATZYQXJA6ATZG","short_pith_number":"pith:M3NGTFXY","canonical_record":{"source":{"id":"2505.19184","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-25T15:06:17Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"ffc3e418e0da706db5c3403f798b318e56526bbb3e2144d598b2fa5f7f12b072","abstract_canon_sha256":"5503442ec8badde93c18bff1d7a03d32946e7525f2a5909d51091febfb66bf53"},"schema_version":"1.0"},"canonical_sha256":"66da6996f86e299d0279c42e907813c984be6a241847349da930c9a7718419db","source":{"kind":"arxiv","id":"2505.19184","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.19184","created_at":"2026-07-05T11:18:22Z"},{"alias_kind":"arxiv_version","alias_value":"2505.19184v3","created_at":"2026-07-05T11:18:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.19184","created_at":"2026-07-05T11:18:22Z"},{"alias_kind":"pith_short_12","alias_value":"M3NGTFXYNYUZ","created_at":"2026-07-05T11:18:22Z"},{"alias_kind":"pith_short_16","alias_value":"M3NGTFXYNYUZ2ATZ","created_at":"2026-07-05T11:18:22Z"},{"alias_kind":"pith_short_8","alias_value":"M3NGTFXY","created_at":"2026-07-05T11:18:22Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:M3NGTFXYNYUZ2ATZYQXJA6ATZG","target":"record","payload":{"canonical_record":{"source":{"id":"2505.19184","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-25T15:06:17Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"ffc3e418e0da706db5c3403f798b318e56526bbb3e2144d598b2fa5f7f12b072","abstract_canon_sha256":"5503442ec8badde93c18bff1d7a03d32946e7525f2a5909d51091febfb66bf53"},"schema_version":"1.0"},"canonical_sha256":"66da6996f86e299d0279c42e907813c984be6a241847349da930c9a7718419db","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:18:22.134984Z","signature_b64":"6dFTHVW2t9pC6HSvt2S2RPCND50oN+WmEE59kcPRK65xPy3Hr4kkS/7pGBLJX7nDePZiAAnC/rY1s9OpEkHpAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"66da6996f86e299d0279c42e907813c984be6a241847349da930c9a7718419db","last_reissued_at":"2026-07-05T11:18:22.134518Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:18:22.134518Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.19184","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":"2026-07-05T11:18:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kzRYG8MU4DzRYAx9sN//oEwN4+Ja10ssmhZBpXtN+WnwBdqjoDuS+EMSRuxDRLu0zQFCly4+3ehBIIxbSfgxBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T06:58:50.720448Z"},"content_sha256":"91720ad87ac0986e5b605ac10784329ff6a5d6ff23e8608f018f959038b0bdb0","schema_version":"1.0","event_id":"sha256:91720ad87ac0986e5b605ac10784329ff6a5d6ff23e8608f018f959038b0bdb0"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:M3NGTFXYNYUZ2ATZYQXJA6ATZG","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"When Two LLMs Debate, Both Think They'll Win","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Minh Nhat Nguyen, Pradyumna Shyama Prasad","submitted_at":"2025-05-25T15:06:17Z","abstract_excerpt":"Can LLMs accurately adjust their confidence when facing opposition? Building on previous studies measuring calibration on static fact-based question-answering tasks, we evaluate Large Language Models (LLMs) in a dynamic, adversarial debate setting, uniquely combining two realistic factors: (a) a multi-turn format requiring models to update beliefs as new information emerges, and (b) a zero-sum structure to control for task-related uncertainty, since mutual high-confidence claims imply systematic overconfidence. We organized 60 three-round policy debates among ten state-of-the-art LLMs, with mo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.19184","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/2505.19184/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":"2026-07-05T11:18:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"S3v5oyvVb5ZrrXfsHqS/sRg5hiTYqg5cI58BRai/yvzRHfYhrB+gepWhy2VQSnlBhpVwufDLmM3D/Q2x8HrzCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T06:58:50.720833Z"},"content_sha256":"43b7d4a8217d16eb360a66c2e4492500c6aaf1f2f275b2b9c5fbf54dd28108da","schema_version":"1.0","event_id":"sha256:43b7d4a8217d16eb360a66c2e4492500c6aaf1f2f275b2b9c5fbf54dd28108da"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/M3NGTFXYNYUZ2ATZYQXJA6ATZG/bundle.json","state_url":"https://pith.science/pith/M3NGTFXYNYUZ2ATZYQXJA6ATZG/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/M3NGTFXYNYUZ2ATZYQXJA6ATZG/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":"2026-07-06T06:58:50Z","links":{"resolver":"https://pith.science/pith/M3NGTFXYNYUZ2ATZYQXJA6ATZG","bundle":"https://pith.science/pith/M3NGTFXYNYUZ2ATZYQXJA6ATZG/bundle.json","state":"https://pith.science/pith/M3NGTFXYNYUZ2ATZYQXJA6ATZG/state.json","well_known_bundle":"https://pith.science/.well-known/pith/M3NGTFXYNYUZ2ATZYQXJA6ATZG/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:M3NGTFXYNYUZ2ATZYQXJA6ATZG","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":"5503442ec8badde93c18bff1d7a03d32946e7525f2a5909d51091febfb66bf53","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-25T15:06:17Z","title_canon_sha256":"ffc3e418e0da706db5c3403f798b318e56526bbb3e2144d598b2fa5f7f12b072"},"schema_version":"1.0","source":{"id":"2505.19184","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.19184","created_at":"2026-07-05T11:18:22Z"},{"alias_kind":"arxiv_version","alias_value":"2505.19184v3","created_at":"2026-07-05T11:18:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.19184","created_at":"2026-07-05T11:18:22Z"},{"alias_kind":"pith_short_12","alias_value":"M3NGTFXYNYUZ","created_at":"2026-07-05T11:18:22Z"},{"alias_kind":"pith_short_16","alias_value":"M3NGTFXYNYUZ2ATZ","created_at":"2026-07-05T11:18:22Z"},{"alias_kind":"pith_short_8","alias_value":"M3NGTFXY","created_at":"2026-07-05T11:18:22Z"}],"graph_snapshots":[{"event_id":"sha256:43b7d4a8217d16eb360a66c2e4492500c6aaf1f2f275b2b9c5fbf54dd28108da","target":"graph","created_at":"2026-07-05T11:18:22Z","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/2505.19184/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Can LLMs accurately adjust their confidence when facing opposition? Building on previous studies measuring calibration on static fact-based question-answering tasks, we evaluate Large Language Models (LLMs) in a dynamic, adversarial debate setting, uniquely combining two realistic factors: (a) a multi-turn format requiring models to update beliefs as new information emerges, and (b) a zero-sum structure to control for task-related uncertainty, since mutual high-confidence claims imply systematic overconfidence. We organized 60 three-round policy debates among ten state-of-the-art LLMs, with mo","authors_text":"Minh Nhat Nguyen, Pradyumna Shyama Prasad","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-25T15:06:17Z","title":"When Two LLMs Debate, Both Think They'll Win"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.19184","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":"sha256:91720ad87ac0986e5b605ac10784329ff6a5d6ff23e8608f018f959038b0bdb0","target":"record","created_at":"2026-07-05T11:18:22Z","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":"5503442ec8badde93c18bff1d7a03d32946e7525f2a5909d51091febfb66bf53","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-25T15:06:17Z","title_canon_sha256":"ffc3e418e0da706db5c3403f798b318e56526bbb3e2144d598b2fa5f7f12b072"},"schema_version":"1.0","source":{"id":"2505.19184","kind":"arxiv","version":3}},"canonical_sha256":"66da6996f86e299d0279c42e907813c984be6a241847349da930c9a7718419db","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"66da6996f86e299d0279c42e907813c984be6a241847349da930c9a7718419db","first_computed_at":"2026-07-05T11:18:22.134518Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:18:22.134518Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"6dFTHVW2t9pC6HSvt2S2RPCND50oN+WmEE59kcPRK65xPy3Hr4kkS/7pGBLJX7nDePZiAAnC/rY1s9OpEkHpAw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:18:22.134984Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.19184","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:91720ad87ac0986e5b605ac10784329ff6a5d6ff23e8608f018f959038b0bdb0","sha256:43b7d4a8217d16eb360a66c2e4492500c6aaf1f2f275b2b9c5fbf54dd28108da"],"state_sha256":"b24e1033f16952f571d45f6020ffb527aef31cea79aa8436be7f40034f327855"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ohocjAdxh+0b+f/cuJ0/YCVH8VakfMzIv77cFg50p8giH4bOB3qNYBnImjow5rs2CUyhNxFhaiFcB42ssplQAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T06:58:50.722809Z","bundle_sha256":"e424a450fa08b4c57e7a36fba25597c63ce304f0256cdbcb14c8e03d4c016061"}}