{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:ZL43OG6SEVN4YUV2PLCGYKVL7W","short_pith_number":"pith:ZL43OG6S","schema_version":"1.0","canonical_sha256":"caf9b71bd2255bcc52ba7ac46c2aabfd8c4589981c06af40bbf7cdeaed6eefe1","source":{"kind":"arxiv","id":"2605.19193","version":1},"attestation_state":"computed","paper":{"title":"Sequential Consensus for Multi-Agent LLM Debates: A Wald-SPRT compute governor with calibration-based failure detection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Andrea Morandi","submitted_at":"2026-05-18T23:43:12Z","abstract_excerpt":"Multi-agent LLM debate improves factuality and reasoning, but most recipes pick a fixed round count, over-spending on easy items and under-spending on hard ones. We adapt Wald's Sequential Probability Ratio Test (SPRT) as a plug-in compute governor for LLM debates. After each round, an LLM judge emits a [0,1] consensus score on the latest agent positions; a Wald monitor accumulates the log-likelihood ratio of \"useful convergence\" vs \"not yet useful\" under a Beta likelihood family, and stops when either boundary is crossed or returns a capped best-effort outcome at R_max. Under i.i.d. assumptio"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2605.19193","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T23:43:12Z","cross_cats_sorted":[],"title_canon_sha256":"23f5b0b32d210e4e2eb0a174d6f667136cdcaafcb33df60f3edf4940af873834","abstract_canon_sha256":"42a6c6709a12ff493325ae78e971e75fe60ef9bf1144aca50021a6d97517f666"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:05:32.882285Z","signature_b64":"zdwq2U7tF1wL7cwfYdZ4cHQBuN88qNjpc+BhWYwqE4vNn27CHyF+v5pr+MLXwlazfQuQFSJrDfx5YozBLuMqBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"caf9b71bd2255bcc52ba7ac46c2aabfd8c4589981c06af40bbf7cdeaed6eefe1","last_reissued_at":"2026-05-20T01:05:32.881576Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:05:32.881576Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Sequential Consensus for Multi-Agent LLM Debates: A Wald-SPRT compute governor with calibration-based failure detection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Andrea Morandi","submitted_at":"2026-05-18T23:43:12Z","abstract_excerpt":"Multi-agent LLM debate improves factuality and reasoning, but most recipes pick a fixed round count, over-spending on easy items and under-spending on hard ones. We adapt Wald's Sequential Probability Ratio Test (SPRT) as a plug-in compute governor for LLM debates. After each round, an LLM judge emits a [0,1] consensus score on the latest agent positions; a Wald monitor accumulates the log-likelihood ratio of \"useful convergence\" vs \"not yet useful\" under a Beta likelihood family, and stops when either boundary is crossed or returns a capped best-effort outcome at R_max. Under i.i.d. assumptio"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.19193","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/2605.19193/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2605.19193","created_at":"2026-05-20T01:05:32.881696+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.19193v1","created_at":"2026-05-20T01:05:32.881696+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.19193","created_at":"2026-05-20T01:05:32.881696+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZL43OG6SEVN4","created_at":"2026-05-20T01:05:32.881696+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZL43OG6SEVN4YUV2","created_at":"2026-05-20T01:05:32.881696+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZL43OG6S","created_at":"2026-05-20T01:05:32.881696+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZL43OG6SEVN4YUV2PLCGYKVL7W","json":"https://pith.science/pith/ZL43OG6SEVN4YUV2PLCGYKVL7W.json","graph_json":"https://pith.science/api/pith-number/ZL43OG6SEVN4YUV2PLCGYKVL7W/graph.json","events_json":"https://pith.science/api/pith-number/ZL43OG6SEVN4YUV2PLCGYKVL7W/events.json","paper":"https://pith.science/paper/ZL43OG6S"},"agent_actions":{"view_html":"https://pith.science/pith/ZL43OG6SEVN4YUV2PLCGYKVL7W","download_json":"https://pith.science/pith/ZL43OG6SEVN4YUV2PLCGYKVL7W.json","view_paper":"https://pith.science/paper/ZL43OG6S","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.19193&json=true","fetch_graph":"https://pith.science/api/pith-number/ZL43OG6SEVN4YUV2PLCGYKVL7W/graph.json","fetch_events":"https://pith.science/api/pith-number/ZL43OG6SEVN4YUV2PLCGYKVL7W/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZL43OG6SEVN4YUV2PLCGYKVL7W/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZL43OG6SEVN4YUV2PLCGYKVL7W/action/storage_attestation","attest_author":"https://pith.science/pith/ZL43OG6SEVN4YUV2PLCGYKVL7W/action/author_attestation","sign_citation":"https://pith.science/pith/ZL43OG6SEVN4YUV2PLCGYKVL7W/action/citation_signature","submit_replication":"https://pith.science/pith/ZL43OG6SEVN4YUV2PLCGYKVL7W/action/replication_record"}},"created_at":"2026-05-20T01:05:32.881696+00:00","updated_at":"2026-05-20T01:05:32.881696+00:00"}