{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:UA7BKEKIBN7GPJIIWGD2J2IJC6","short_pith_number":"pith:UA7BKEKI","canonical_record":{"source":{"id":"2510.20963","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-23T19:46:00Z","cross_cats_sorted":[],"title_canon_sha256":"21a5eec8b9f68642e1b58a8a9cf57a823d7e8d374d1ebcf426e308172186bbe8","abstract_canon_sha256":"3e455e817156fab3b14103b7fd304bd78c8bed8d8d556258a705a01545258974"},"schema_version":"1.0"},"canonical_sha256":"a03e1511480b7e67a508b187a4e909178b4d5eb9432d76db26cf58c6732eb3e5","source":{"kind":"arxiv","id":"2510.20963","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2510.20963","created_at":"2026-07-15T01:21:47Z"},{"alias_kind":"arxiv_version","alias_value":"2510.20963v2","created_at":"2026-07-15T01:21:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.20963","created_at":"2026-07-15T01:21:47Z"},{"alias_kind":"pith_short_12","alias_value":"UA7BKEKIBN7G","created_at":"2026-07-15T01:21:47Z"},{"alias_kind":"pith_short_16","alias_value":"UA7BKEKIBN7GPJII","created_at":"2026-07-15T01:21:47Z"},{"alias_kind":"pith_short_8","alias_value":"UA7BKEKI","created_at":"2026-07-15T01:21:47Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:UA7BKEKIBN7GPJIIWGD2J2IJC6","target":"record","payload":{"canonical_record":{"source":{"id":"2510.20963","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-23T19:46:00Z","cross_cats_sorted":[],"title_canon_sha256":"21a5eec8b9f68642e1b58a8a9cf57a823d7e8d374d1ebcf426e308172186bbe8","abstract_canon_sha256":"3e455e817156fab3b14103b7fd304bd78c8bed8d8d556258a705a01545258974"},"schema_version":"1.0"},"canonical_sha256":"a03e1511480b7e67a508b187a4e909178b4d5eb9432d76db26cf58c6732eb3e5","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-15T01:21:47.620021Z","signature_b64":"pJ8NXCMVGF2KQ2uhLL/FUqPBC52TnbAXeRjYBpZuYjpR4c0W/Y1MSMteGz3vT2/Xsods4y0ak2w06ktwr0SDAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a03e1511480b7e67a508b187a4e909178b4d5eb9432d76db26cf58c6732eb3e5","last_reissued_at":"2026-07-15T01:21:47.619080Z","signature_status":"signed_v1","first_computed_at":"2026-07-15T01:21:47.619080Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2510.20963","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-15T01:21:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zbDaOmavduPRK/ZYtY+/5xQd+DTAbDYDw61EsDLHLDHG+MckGeDV8xkT2dgkuciX6R3CIxc2P6jTgb4LAEL5BQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T04:57:46.042521Z"},"content_sha256":"d9c804e546a8b04a301ea3227eeec22ae9f9d39a4c6edbd69dd32f55c1185936","schema_version":"1.0","event_id":"sha256:d9c804e546a8b04a301ea3227eeec22ae9f9d39a4c6edbd69dd32f55c1185936"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:UA7BKEKIBN7GPJIIWGD2J2IJC6","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"When and Why Does Multi-Agent Debate Fail and Does It Really Underperform?","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bo Han, Gang Niu, James Cheng, Masashi Sugiyama, Yongqiang Chen","submitted_at":"2025-10-23T19:46:00Z","abstract_excerpt":"Multi-agent debate (MAD) was proposed as a promising approach for ensembling the wisdom of multiple large language models (LLMs) to improve reasoning and provide effective supervision to superhuman LLMs. However, increasing empirical evidence suggests that MAD may not outperform or even significantly underperform single-agent approaches (SA), raising doubts about the benefits of MAD. In this work, we investigate this issue by analyzing the incentive structures of popular MAD paradigms: (i) competitive MAD (CopMAD) where agents compete by holding opposing positions; (ii) consensus-seeking MAD ("},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.20963","kind":"arxiv","version":2},"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/2510.20963/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":"2026-07-15T01:21:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kTkQmaNmn4WmJfhZ/QLBbDUdxmzPMMaaj9EbdYK2KTc4Eqr+ysx00u4xtTNcOwNrBmyaciNUv5jB/nMQuIBwAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T04:57:46.043014Z"},"content_sha256":"5595087b0b495886cc12c3cd8abd87603f63bc3899e8f1345901ff581f4a803e","schema_version":"1.0","event_id":"sha256:5595087b0b495886cc12c3cd8abd87603f63bc3899e8f1345901ff581f4a803e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/UA7BKEKIBN7GPJIIWGD2J2IJC6/bundle.json","state_url":"https://pith.science/pith/UA7BKEKIBN7GPJIIWGD2J2IJC6/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/UA7BKEKIBN7GPJIIWGD2J2IJC6/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":"2026-08-05T04:57:46Z","links":{"resolver":"https://pith.science/pith/UA7BKEKIBN7GPJIIWGD2J2IJC6","bundle":"https://pith.science/pith/UA7BKEKIBN7GPJIIWGD2J2IJC6/bundle.json","state":"https://pith.science/pith/UA7BKEKIBN7GPJIIWGD2J2IJC6/state.json","well_known_bundle":"https://pith.science/.well-known/pith/UA7BKEKIBN7GPJIIWGD2J2IJC6/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:UA7BKEKIBN7GPJIIWGD2J2IJC6","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":"3e455e817156fab3b14103b7fd304bd78c8bed8d8d556258a705a01545258974","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-23T19:46:00Z","title_canon_sha256":"21a5eec8b9f68642e1b58a8a9cf57a823d7e8d374d1ebcf426e308172186bbe8"},"schema_version":"1.0","source":{"id":"2510.20963","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2510.20963","created_at":"2026-07-15T01:21:47Z"},{"alias_kind":"arxiv_version","alias_value":"2510.20963v2","created_at":"2026-07-15T01:21:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.20963","created_at":"2026-07-15T01:21:47Z"},{"alias_kind":"pith_short_12","alias_value":"UA7BKEKIBN7G","created_at":"2026-07-15T01:21:47Z"},{"alias_kind":"pith_short_16","alias_value":"UA7BKEKIBN7GPJII","created_at":"2026-07-15T01:21:47Z"},{"alias_kind":"pith_short_8","alias_value":"UA7BKEKI","created_at":"2026-07-15T01:21:47Z"}],"graph_snapshots":[{"event_id":"sha256:5595087b0b495886cc12c3cd8abd87603f63bc3899e8f1345901ff581f4a803e","target":"graph","created_at":"2026-07-15T01:21:47Z","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/2510.20963/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multi-agent debate (MAD) was proposed as a promising approach for ensembling the wisdom of multiple large language models (LLMs) to improve reasoning and provide effective supervision to superhuman LLMs. However, increasing empirical evidence suggests that MAD may not outperform or even significantly underperform single-agent approaches (SA), raising doubts about the benefits of MAD. In this work, we investigate this issue by analyzing the incentive structures of popular MAD paradigms: (i) competitive MAD (CopMAD) where agents compete by holding opposing positions; (ii) consensus-seeking MAD (","authors_text":"Bo Han, Gang Niu, James Cheng, Masashi Sugiyama, Yongqiang Chen","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-23T19:46:00Z","title":"When and Why Does Multi-Agent Debate Fail and Does It Really Underperform?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.20963","kind":"arxiv","version":2},"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:d9c804e546a8b04a301ea3227eeec22ae9f9d39a4c6edbd69dd32f55c1185936","target":"record","created_at":"2026-07-15T01:21:47Z","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":"3e455e817156fab3b14103b7fd304bd78c8bed8d8d556258a705a01545258974","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-23T19:46:00Z","title_canon_sha256":"21a5eec8b9f68642e1b58a8a9cf57a823d7e8d374d1ebcf426e308172186bbe8"},"schema_version":"1.0","source":{"id":"2510.20963","kind":"arxiv","version":2}},"canonical_sha256":"a03e1511480b7e67a508b187a4e909178b4d5eb9432d76db26cf58c6732eb3e5","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a03e1511480b7e67a508b187a4e909178b4d5eb9432d76db26cf58c6732eb3e5","first_computed_at":"2026-07-15T01:21:47.619080Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-15T01:21:47.619080Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"pJ8NXCMVGF2KQ2uhLL/FUqPBC52TnbAXeRjYBpZuYjpR4c0W/Y1MSMteGz3vT2/Xsods4y0ak2w06ktwr0SDAg==","signature_status":"signed_v1","signed_at":"2026-07-15T01:21:47.620021Z","signed_message":"canonical_sha256_bytes"},"source_id":"2510.20963","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d9c804e546a8b04a301ea3227eeec22ae9f9d39a4c6edbd69dd32f55c1185936","sha256:5595087b0b495886cc12c3cd8abd87603f63bc3899e8f1345901ff581f4a803e"],"state_sha256":"5123ee9e93420fa0e806647c5f770efb625cbb6fdf4405f110b6c467b71f194d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"O+/GSBKOg+2StFF8MtW4Nu03xXn1ZLDBnkoLZOAsu2iBx5Xf7wI+RuTutJMNDg1s1it6HYzHPqhirJtrSuvNBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T04:57:46.048083Z","bundle_sha256":"7be1a6f51dfbba0a0f3cf735c0a4f68f9bfe39de2697e73905212149ecbd878b"}}