Optimal Aggregation of LLM and PRM Signals for Efficient Test-Time Scaling Article Swipe

MAKE IMPACT

Yanli Wang , Xiufeng Han , Ke Xu , Haohan Wang · YOU? · · 2025 · Open Access · · DOI: https://doi.org/10.48550/arxiv.2510.13918

Process reward models (PRMs) are a cornerstone of test-time scaling (TTS), designed to verify and select the best responses from large language models (LLMs). However, this promise is challenged by recent benchmarks where simple majority voting, which ignores PRM signals, occasionally outperforms standard PRM-based selection. This raises a critical question: How can we effectively utilize verification signals from PRMs for TTS? To address this, we start by developing a theoretical framework for optimally combining signals from both the LLM and the PRM. Our framework reveals that the optimal strategy is a weighted aggregation of responses, a strategy whose effectiveness hinges on estimating weights that capture the complex interplay between the models. Based on our theoretical results, we empirically show that these optimal weighting functions differ significantly across LLM-PRM pairs and, notably, often assign substantial negative weights. Motivated by these insights, we propose efficient pre-computation methods to calibrate these weighting functions. Extensive experiments across 5 LLMs and 7 PRMs demonstrate that our calibration method significantly boosts the TTS efficiency, surpassing the performance of vanilla weighted majority voting while using only $21.3\%$ of the computation. Ultimately, our work demonstrates that investing in a more intelligent aggregation strategy can be a more convincing path to performance gains than simply scaling test-time computation.

Type

preprint

Landing Page

http://arxiv.org/abs/2510.13918

PDF

https://arxiv.org/pdf/2510.13918

OA Status

green

OpenAlex ID

https://openalex.org/W4416145571

Raw OpenAlex JSON

OpenAlex ID

https://openalex.org/W4416145571

Canonical identifier for this work in OpenAlex

DOI

https://doi.org/10.48550/arxiv.2510.13918

Digital Object Identifier

Title

Optimal Aggregation of LLM and PRM Signals for Efficient Test-Time Scaling

Work title

Type

preprint

OpenAlex work type

Publication year

2025

Year of publication

Publication date

2025-10-15

Full publication date if available

Authors

Yanli Wang, Xiufeng Han, Ke Xu, Haohan Wang

List of authors in order

Landing page

https://arxiv.org/abs/2510.13918

Publisher landing page

PDF URL

https://arxiv.org/pdf/2510.13918

Direct link to full text PDF

Open access

Yes

Whether a free full text is available

OA status

green

Open access status per OpenAlex

OA URL

https://arxiv.org/pdf/2510.13918

Direct OA link when available

Cited by

0

Total citation count in OpenAlex

Full payload