Budget-aware Test-time Scaling via Discriminative Verification

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Montgomery, Kyle, Tan, Sijun, Chen, Yuqi, Zhuang, Siyuan, Zhang, Tianjun, Popa, Raluca Ada, Wang, Chenguang
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914096655368192
author Montgomery, Kyle
Tan, Sijun
Chen, Yuqi
Zhuang, Siyuan
Zhang, Tianjun
Popa, Raluca Ada
Wang, Chenguang
author_facet Montgomery, Kyle
Tan, Sijun
Chen, Yuqi
Zhuang, Siyuan
Zhang, Tianjun
Popa, Raluca Ada
Wang, Chenguang
contents Test-time scaling is a powerful strategy for boosting the performance of large language models on complex reasoning tasks. While state-of-the-art approaches often employ generative verifiers to select the best solution from a pool of candidates, this method incurs prohibitive computational costs, limiting its practicality. In this work, we shift the focus to a more budget-aware paradigm: discriminative verification. We conduct a thorough empirical analysis and demonstrate that while discriminative verifiers may underperform in isolation, combining them with self-consistency in a hybrid approach creates a powerful and efficient test-time scaling mechanism. Notably, under a fixed compute budget, this hybrid approach surpasses state-of-the-art generative verification by a significant margin: achieving up to 15.3\% higher accuracy on AIME2025. Our findings establish that for practical, real-world applications, budget-aware scaling with discriminative verifiers is not only a "free" upgrade over self-consistency, but also a more effective and efficient alternative to costly generative techniques. Code is available at https://github.com/wang-research-lab/verification.
format Preprint
id arxiv_https___arxiv_org_abs_2510_14913
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Budget-aware Test-time Scaling via Discriminative Verification
Montgomery, Kyle
Tan, Sijun
Chen, Yuqi
Zhuang, Siyuan
Zhang, Tianjun
Popa, Raluca Ada
Wang, Chenguang
Artificial Intelligence
Computation and Language
Machine Learning
Test-time scaling is a powerful strategy for boosting the performance of large language models on complex reasoning tasks. While state-of-the-art approaches often employ generative verifiers to select the best solution from a pool of candidates, this method incurs prohibitive computational costs, limiting its practicality. In this work, we shift the focus to a more budget-aware paradigm: discriminative verification. We conduct a thorough empirical analysis and demonstrate that while discriminative verifiers may underperform in isolation, combining them with self-consistency in a hybrid approach creates a powerful and efficient test-time scaling mechanism. Notably, under a fixed compute budget, this hybrid approach surpasses state-of-the-art generative verification by a significant margin: achieving up to 15.3\% higher accuracy on AIME2025. Our findings establish that for practical, real-world applications, budget-aware scaling with discriminative verifiers is not only a "free" upgrade over self-consistency, but also a more effective and efficient alternative to costly generative techniques. Code is available at https://github.com/wang-research-lab/verification.
title Budget-aware Test-time Scaling via Discriminative Verification
topic Artificial Intelligence
Computation and Language
Machine Learning
url https://arxiv.org/abs/2510.14913