Saved in:
Bibliographic Details
Main Authors: Jing, Jinhao, Zhao, Qiannian, Huang, Chao, Su, Zhan
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.07488
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911662006599680
author Jing, Jinhao
Zhao, Qiannian
Huang, Chao
Su, Zhan
author_facet Jing, Jinhao
Zhao, Qiannian
Huang, Chao
Su, Zhan
contents The scaling of Large Multimodal Models (LMMs) is constrained by the quality-quantity trade-off inherent in synthetic data. Previous approaches, such as LLM-as-a-Judge, have proven their effectiveness in addressing this but suffer from prohibitive computational costs and lack of interpretability. To bridge this gap, we propose One-Step-Train (OST), a framework that reformulates data selection as an incremental optimization utility ranking problem. Instead of relying on semantic heuristics, OST estimates the marginal utility of each sample via a simulated single-step update on a lightweight proxy. Experiments on the Qwen series across multimodal mathematical reasoning benchmarks demonstrate that OST achieves Pareto-optimal efficiency. By selecting the top-50 subset, OST reduces training costs by 43% (and total time consumption by 17) while surpassing the strong LLM-as-a-Judge baseline by 1.8 points. Furthermore, under a fixed compute budget, our method using only the top-20 subset achieves a 5.6 point gain over LLM-as-a-Judge, improves upon heuristic scoring baselines like DEITA, and outperforms the Full-SFT baseline by 8.8 points. Notably, while Full-SFT suffers from performance degradation due to noise, our optimization-grounded approach effectively identifies toxic samples, successfully reversing the negative transfer frequently observed in complex reasoning tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2605_07488
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Efficient Data Selection for Multimodal Models via Incremental Optimization Utility
Jing, Jinhao
Zhao, Qiannian
Huang, Chao
Su, Zhan
Artificial Intelligence
Machine Learning
The scaling of Large Multimodal Models (LMMs) is constrained by the quality-quantity trade-off inherent in synthetic data. Previous approaches, such as LLM-as-a-Judge, have proven their effectiveness in addressing this but suffer from prohibitive computational costs and lack of interpretability. To bridge this gap, we propose One-Step-Train (OST), a framework that reformulates data selection as an incremental optimization utility ranking problem. Instead of relying on semantic heuristics, OST estimates the marginal utility of each sample via a simulated single-step update on a lightweight proxy. Experiments on the Qwen series across multimodal mathematical reasoning benchmarks demonstrate that OST achieves Pareto-optimal efficiency. By selecting the top-50 subset, OST reduces training costs by 43% (and total time consumption by 17) while surpassing the strong LLM-as-a-Judge baseline by 1.8 points. Furthermore, under a fixed compute budget, our method using only the top-20 subset achieves a 5.6 point gain over LLM-as-a-Judge, improves upon heuristic scoring baselines like DEITA, and outperforms the Full-SFT baseline by 8.8 points. Notably, while Full-SFT suffers from performance degradation due to noise, our optimization-grounded approach effectively identifies toxic samples, successfully reversing the negative transfer frequently observed in complex reasoning tasks.
title Efficient Data Selection for Multimodal Models via Incremental Optimization Utility
topic Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2605.07488