Self-Consistency from Only Two Samples: CoT-PoT Ensembling for Efficient LLM Reasoning

arXiv:2604.17433v1 Announce Type: new Abstract: 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 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.

Related

• STU-PID: Steering Token Usage via PID Controller for Efficient Large Language Model Reasoning
• Test-Time Reasoners Are Strategic Multiple-Choice Test-Takers
• ThinkBrake: Efficient Reasoning via Log-Probability Margin Guided Decoding
• DIVERSED: Relaxed Speculative Decoding via Dynamic Ensemble Verification
• Process Reward Models Meet Planning: Generating Precise and Scalable Datasets for Step-Level Rewards

Source: arXiv cs.CL | 2026-04-21