{"paper":{"title":"Self-Consistency from Only Two Samples: CoT-PoT Ensembling for Efficient LLM Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"CoT-PoT ensembling cuts the samples needed for LLM self-consistency by 9.3 times while raising accuracy.","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Majd Hawasly, Md Rizwan Parvez, Mohammad Raza, Raman Saparkhan","submitted_at":"2026-04-19T13:26:04Z","abstract_excerpt":"Self-consistency (SC) is a popular technique for improving the reasoning accuracy of large language models by aggregating multiple sampled outputs, but it comes at a high computational cost due to extensive sampling. We introduce a hybrid ensembling approach that leverages the complementary strengths of two distinct modes of reasoning: Chain-of-Thought (CoT) and Program-of-Thought (PoT). We describe a general framework for combining these two forms of reasoning in self-consistency, as well as particular strategies for both full sampling and early-stopping. We show that CoT-PoT ensembling not o"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"CoT-PoT ensembling not only improves overall accuracy, but also drastically reduces the number of samples required for SC by a factor of 9.3x. In particular, the majority of tasks (78.6%) can be addressed with only two samples, which has not been possible with any prior SC methods.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumption that Chain-of-Thought and Program-of-Thought outputs are sufficiently complementary and that their agreement reliably indicates correctness without needing many more samples or introducing new error modes; this is implicit in the early-stopping and ensembling strategies described.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"CoT-PoT ensembling achieves self-consistency accuracy in LLMs with only two samples for 78.6% of tasks, reducing computation by 9.3x compared to standard methods.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"CoT-PoT ensembling cuts the samples needed for LLM self-consistency by 9.3 times while raising accuracy.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"c6e0dcf4bda387a986baee7ec0d0d37fbc37dec72d2195d67574b9a5da52d385"},"source":{"id":"2604.17433","kind":"arxiv","version":2},"verdict":{"id":"feb6b808-20e4-43f4-858c-ca715a8382db","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T05:36:51.832748Z","strongest_claim":"CoT-PoT ensembling not only improves overall accuracy, but also drastically reduces the number of samples required for SC by a factor of 9.3x. In particular, the majority of tasks (78.6%) can be addressed with only two samples, which has not been possible with any prior SC methods.","one_line_summary":"CoT-PoT ensembling achieves self-consistency accuracy in LLMs with only two samples for 78.6% of tasks, reducing computation by 9.3x compared to standard methods.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumption that Chain-of-Thought and Program-of-Thought outputs are sufficiently complementary and that their agreement reliably indicates correctness without needing many more samples or introducing new error modes; this is implicit in the early-stopping and ensembling strategies described.","pith_extraction_headline":"CoT-PoT ensembling cuts the samples needed for LLM self-consistency by 9.3 times while raising accuracy."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.17433/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"}