paper-with-me

홈 › Papers

On Representation Redundancy in Large-Scale Instruction Tuning Data Selection

2026-02-14 · Youwei Shu, Shaomian Zheng, Dingnan Jin, Wenjie Qu, Ziyao Guo, Qing Cui, Jun Zhou, Jiaheng Zhang arxiv

Data quality is a crucial factor in large language models training. While prior work has shown that models trained on smaller, high-quality datasets can outperform those trained on much larger but noisy or low-quality corpora, systematic methods for industrial-scale data selection in instruction tuning remain underexplored. In this work, we study instruction-tuning data selection through the lens of semantic representation similarity and identify a key limitation of state-of-the-art LLM encoders: they produce highly redundant semantic embeddings. To mitigate this redundancy, we propose Compressed Representation Data Selection (CRDS), a novel framework with two variants. CRDS-R applies Rademacher random projection followed by concatenation of transformer hidden-layer representations, while CRDS-W employs whitening-based dimensionality reduction to improve representational quality. Experimental results demonstrate that both variants substantially enhance data quality and consistently outperform state-of-the-art representation-based selection methods. Notably, CRDS-W achieves strong performance using only 3.5% of the data, surpassing the full-data baseline by an average of 0.71% across four datasets. Our code is available at https://github.com/tdano1/CRDS.

📄 PDF Abstract BibTeX arXiv:2602.13773

Code (0)

등록된 구현이 없습니다.

Tasks

Dimensionality Reduction

Similar Papers 제목 키워드 기반

ScalSelect: Scalable Training-Free Multimodal Data Selection for Efficient Visual Instruction Tuning

2026-02-12 · Changti Wu, Jiahuai Mao, Yuzhuo Miao, Shijie Lian 외 arxiv

Large-scale Visual Instruction Tuning (VIT) has become a key paradigm for advancing the performance of vision-language models (VLMs) across various multimodal tasks. However, training on the large-scale datasets is compu…

Less is More: High-value Data Selection for Visual Instruction Tuning

2024-03-14 · Zikang Liu, Kun Zhou, Wayne Xin Zhao, Dawei Gao 외

Visual instruction tuning is the key to building large vision language models~(LVLMs), which can greatly improve the task generalization and solving capabilities by learning a mixture of instruction data from diverse vis…

Data Selection for Fine-tuning Vision Language Models via Cross Modal Alignment Trajectories

2025-10-01 · Nilay Naharas, Dang Nguyen, Nesihan Bulut, Mohammadhossein Bateni 외 arxiv

Data-efficient learning aims to eliminate redundancy in large training datasets by training models on smaller subsets of the most informative examples. While data selection has been extensively explored for vision models…

Exploring Instruction Data Quality for Explainable Image Quality Assessment

2025-10-04 · Yunhao Li, Sijing Wu, Huiyu Duan, Yucheng Zhu 외 arxiv

In recent years, with the rapid development of large multimodal models (LMMs), explainable image quality assessment (IQA) has attracted increasing attention, aiming to understand the perceptual quality problems of images…

Image Quality Assessment

Federated Data-Efficient Instruction Tuning for Large Language Models

2024-10-14 · Zhen Qin, Zhaomin Wu, Bingsheng He, Shuiguang Deng

Instruction tuning helps improve pretrained large language models (LLMs) in terms of the responsiveness to human instructions, which is benefited from diversified instruction data. Federated learning extends the sources …

Federated Learning