Strategic Self-Consistency
Abstract
Self-consistency has become a popular technique for enhancing the reasoning abilities of large language models by generating multiple reasoning paths and selecting the final answer through a majority vote. However, because model providers typically charge users in proportion to the number of reasoning paths generated, they have a financial incentive to artificially increase the path count. In this work, we show that an unfaithful provider can exploit this incentive using a simple, efficient algorithm while avoiding detection by an auditor: by generating and strategically reordering additional reasoning paths, the algorithm makes every path appear necessary to reach the majority. To validate our algorithm, we conduct experiments with multiple instruct models from the Llama and Qwen families, as well as reasoning models distilled from DeepSeek-R1, on benchmark datasets spanning mathematics, science, and question answering. Our results suggest that the distribution of additional reasoning paths generated by our algorithm is heavy-tailed and that substantial capacity to overcharge remains even under the best possible audit designed to keep the false-positive rate below α= 0.1.
Related
- Strategic Interview Positioning under Uncertain Self-Rank
- Self-Consistency via Marginal Sharpening
- Strong Consistency, Optimal Tracking and Sharp Rates for the Åström-Wittenmark Self-Tuning Regulator with Unknown Input Gain
- Self-consistency of the chiral soliton: the heat kernel constrains the Lagrangian
- The Pathology of the Continuum and the Discrete Substrate:The Inevitable Requirement for Mathematical Self-Consistency
BibTeX
@misc{qiu2026strategic,
title = {{Strategic Self-Consistency}},
author = {Tori Qiu and Ander Artola Velasco and Manuel Gomez-Rodriguez},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.30352},
eprint = {2609.30352},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.30352},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Strategic Self-Consistency
AU - Tori Qiu
AU - Ander Artola Velasco
AU - Manuel Gomez-Rodriguez
PY - 2026
DA - 2026/09/24
PB - arXiv
DO - 10.48550/arxiv.2609.30352
UR - https://arxiv.org/abs/2609.30352
AB - Self-consistency has become a popular technique for enhancing the reasoning abilities of large language models by generating multiple reasoning paths and selecting the final answer through a majority vote. However, because model providers typically charge users in proportion to the number of reasoning paths generated, they have a financial incentive to artificially increase the path count. In this work, we show that an unfaithful provider can exploit this incentive using a simple, efficient algorithm while avoiding detection by an auditor: by generating and strategically reordering additional reasoning paths, the algorithm makes every path appear necessary to reach the majority. To validate our algorithm, we conduct experiments with multiple instruct models from the Llama and Qwen families, as well as reasoning models distilled from DeepSeek-R1, on benchmark datasets spanning mathematics, science, and question answering. Our results suggest that the distribution of additional reasoning paths generated by our algorithm is heavy-tailed and that substantial capacity to overcharge remains even under the best possible audit designed to keep the false-positive rate below α= 0.1.
ER -