Saved in:
Bibliographic Details
Main Authors: Tarunokusumo, Ravindra Aribowo, Cunha, Rafael Fernandes
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.21398
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918169580404736
author Tarunokusumo, Ravindra Aribowo
Cunha, Rafael Fernandes
author_facet Tarunokusumo, Ravindra Aribowo
Cunha, Rafael Fernandes
contents Test-time scaling methods have seen a rapid increase in popularity for its computational efficiency and parameter-independent training to improve reasoning performance on Large Language Models. One such method is called budget forcing, a decoding intervention strategy which allocates extra compute budget for thinking and elicits the inherent self-correcting behavior of the model. However, this relies on supervised fine-tuning (SFT) on long-context reasoning traces which causes performance degradation on smaller models due to verbose responses. For this reason, we offer a framework integrating reinforcement learning (RL) to improve token efficiency and boost the performance of a 1.5B model for mathematical reasoning. We demonstrate this using only 1.5K training samples and found that our SFT+RL model performed better on the GSM8K dataset with varying compute budgets. Our main findings showed an overall higher accuracy while significantly reducing its token usage by over 40% compared to the SFT model, revealing how RL can recover the losses due to long-context training and altogether improving performance in mathematical reasoning.
format Preprint
id arxiv_https___arxiv_org_abs_2510_21398
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Boosting Accuracy and Efficiency of Budget Forcing in LLMs via Reinforcement Learning for Mathematical Reasoning
Tarunokusumo, Ravindra Aribowo
Cunha, Rafael Fernandes
Artificial Intelligence
Test-time scaling methods have seen a rapid increase in popularity for its computational efficiency and parameter-independent training to improve reasoning performance on Large Language Models. One such method is called budget forcing, a decoding intervention strategy which allocates extra compute budget for thinking and elicits the inherent self-correcting behavior of the model. However, this relies on supervised fine-tuning (SFT) on long-context reasoning traces which causes performance degradation on smaller models due to verbose responses. For this reason, we offer a framework integrating reinforcement learning (RL) to improve token efficiency and boost the performance of a 1.5B model for mathematical reasoning. We demonstrate this using only 1.5K training samples and found that our SFT+RL model performed better on the GSM8K dataset with varying compute budgets. Our main findings showed an overall higher accuracy while significantly reducing its token usage by over 40% compared to the SFT model, revealing how RL can recover the losses due to long-context training and altogether improving performance in mathematical reasoning.
title Boosting Accuracy and Efficiency of Budget Forcing in LLMs via Reinforcement Learning for Mathematical Reasoning
topic Artificial Intelligence
url https://arxiv.org/abs/2510.21398