Think, Prune, Train, Improve: Scaling Reasoning without Scaling Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Costello, Caia, Guo, Simon, Goldie, Anna, Mirhoseini, Azalia
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915258433536000
author Costello, Caia
Guo, Simon
Goldie, Anna
Mirhoseini, Azalia
author_facet Costello, Caia
Guo, Simon
Goldie, Anna
Mirhoseini, Azalia
contents Large language models (LLMs) have demonstrated strong capabilities in programming and mathematical reasoning tasks, but are constrained by limited high-quality training data. Synthetic data can be leveraged to enhance fine-tuning outcomes, but several factors influence this process, including model size, synthetic data volume, pruning strategy, and number of fine-tuning rounds. We explore these axes and investigate which conditions enable model self-improvement. We introduce the Think, Prune, Train process, a scalable framework that iteratively fine-tunes models on their own reasoning traces, using ground-truth pruning to ensure high-quality training data. This approach yields improved performance: on GSM8K, Gemma2-2B achieves a Pass@1 of 57.6% (from 41.9%), Gemma2-9B reaches 82%, matching LLaMA-3.1-70B, and LLaMA-3.1-70B attains 91%, even surpassing GPT-4o, demonstrating the effectiveness of self-generated reasoning and systematic data selection for improving LLM capabilities.
format Preprint
id arxiv_https___arxiv_org_abs_2504_18116
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Think, Prune, Train, Improve: Scaling Reasoning without Scaling Models
Costello, Caia
Guo, Simon
Goldie, Anna
Mirhoseini, Azalia
Machine Learning
Large language models (LLMs) have demonstrated strong capabilities in programming and mathematical reasoning tasks, but are constrained by limited high-quality training data. Synthetic data can be leveraged to enhance fine-tuning outcomes, but several factors influence this process, including model size, synthetic data volume, pruning strategy, and number of fine-tuning rounds. We explore these axes and investigate which conditions enable model self-improvement. We introduce the Think, Prune, Train process, a scalable framework that iteratively fine-tunes models on their own reasoning traces, using ground-truth pruning to ensure high-quality training data. This approach yields improved performance: on GSM8K, Gemma2-2B achieves a Pass@1 of 57.6% (from 41.9%), Gemma2-9B reaches 82%, matching LLaMA-3.1-70B, and LLaMA-3.1-70B attains 91%, even surpassing GPT-4o, demonstrating the effectiveness of self-generated reasoning and systematic data selection for improving LLM capabilities.
title Think, Prune, Train, Improve: Scaling Reasoning without Scaling Models
topic Machine Learning
url https://arxiv.org/abs/2504.18116