{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:EULKGU3J7VIR3J3GWN36O4VWNK","short_pith_number":"pith:EULKGU3J","schema_version":"1.0","canonical_sha256":"2516a35369fd511da766b377e772b66ab28aedec8c0e5c59e4996a6228f1b989","source":{"kind":"arxiv","id":"2511.02603","version":2},"attestation_state":"computed","paper":{"title":"CGES: Confidence-Guided Early Stopping for Efficient and Accurate Self-Consistency","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ahmad Ghasemi, Ehsan Aghazadeh, Hedyeh Beyhaghi, Hossein Pishro-Nik","submitted_at":"2025-11-04T14:25:54Z","abstract_excerpt":"Large language models (LLMs) are often queried multiple times at test time, with predictions aggregated by majority vote. While effective, this self-consistency (Wang et al., 2023) strategy requires a fixed number of calls and fails when the correct answer is infrequent. We introduce Confidence-Guided Early Stopping (CGES), a Bayesian framework that forms posteriors over candidate answers and adaptively halts sampling once one answer accumulates enough posterior mass. We prove guarantees in both an ideal calibrated regime and a realistic noisy-confidence regime under a directional drift condit"},"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":"2511.02603","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-11-04T14:25:54Z","cross_cats_sorted":[],"title_canon_sha256":"5de0f6b7315881e6f137f99b140a08b7ffcdb85fe729f0b2d93b46206442bb92","abstract_canon_sha256":"2905d1461b29c467edfa0579f29acbb41279a6e566e5bd6c08b7a8eae468a21d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-10T01:09:46.263182Z","signature_b64":"VH3ZqyT6oPMZjVyHB2aHuFA3TBCN4lcmMnp+nM6RzKA4zzwR7NtZN6loKGTBqbhnITUYq1lZBMMfNbOekByVBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2516a35369fd511da766b377e772b66ab28aedec8c0e5c59e4996a6228f1b989","last_reissued_at":"2026-06-10T01:09:46.262430Z","signature_status":"signed_v1","first_computed_at":"2026-06-10T01:09:46.262430Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CGES: Confidence-Guided Early Stopping for Efficient and Accurate Self-Consistency","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ahmad Ghasemi, Ehsan Aghazadeh, Hedyeh Beyhaghi, Hossein Pishro-Nik","submitted_at":"2025-11-04T14:25:54Z","abstract_excerpt":"Large language models (LLMs) are often queried multiple times at test time, with predictions aggregated by majority vote. While effective, this self-consistency (Wang et al., 2023) strategy requires a fixed number of calls and fails when the correct answer is infrequent. We introduce Confidence-Guided Early Stopping (CGES), a Bayesian framework that forms posteriors over candidate answers and adaptively halts sampling once one answer accumulates enough posterior mass. We prove guarantees in both an ideal calibrated regime and a realistic noisy-confidence regime under a directional drift condit"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2511.02603","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/2511.02603/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":"2511.02603","created_at":"2026-06-10T01:09:46.262540+00:00"},{"alias_kind":"arxiv_version","alias_value":"2511.02603v2","created_at":"2026-06-10T01:09:46.262540+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.02603","created_at":"2026-06-10T01:09:46.262540+00:00"},{"alias_kind":"pith_short_12","alias_value":"EULKGU3J7VIR","created_at":"2026-06-10T01:09:46.262540+00:00"},{"alias_kind":"pith_short_16","alias_value":"EULKGU3J7VIR3J3G","created_at":"2026-06-10T01:09:46.262540+00:00"},{"alias_kind":"pith_short_8","alias_value":"EULKGU3J","created_at":"2026-06-10T01:09:46.262540+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.05873","citing_title":"CITE: Anytime-Valid Statistical Inference in LLM Self-Consistency","ref_index":3,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EULKGU3J7VIR3J3GWN36O4VWNK","json":"https://pith.science/pith/EULKGU3J7VIR3J3GWN36O4VWNK.json","graph_json":"https://pith.science/api/pith-number/EULKGU3J7VIR3J3GWN36O4VWNK/graph.json","events_json":"https://pith.science/api/pith-number/EULKGU3J7VIR3J3GWN36O4VWNK/events.json","paper":"https://pith.science/paper/EULKGU3J"},"agent_actions":{"view_html":"https://pith.science/pith/EULKGU3J7VIR3J3GWN36O4VWNK","download_json":"https://pith.science/pith/EULKGU3J7VIR3J3GWN36O4VWNK.json","view_paper":"https://pith.science/paper/EULKGU3J","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2511.02603&json=true","fetch_graph":"https://pith.science/api/pith-number/EULKGU3J7VIR3J3GWN36O4VWNK/graph.json","fetch_events":"https://pith.science/api/pith-number/EULKGU3J7VIR3J3GWN36O4VWNK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EULKGU3J7VIR3J3GWN36O4VWNK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EULKGU3J7VIR3J3GWN36O4VWNK/action/storage_attestation","attest_author":"https://pith.science/pith/EULKGU3J7VIR3J3GWN36O4VWNK/action/author_attestation","sign_citation":"https://pith.science/pith/EULKGU3J7VIR3J3GWN36O4VWNK/action/citation_signature","submit_replication":"https://pith.science/pith/EULKGU3J7VIR3J3GWN36O4VWNK/action/replication_record"}},"created_at":"2026-06-10T01:09:46.262540+00:00","updated_at":"2026-06-10T01:09:46.262540+00:00"}