Neuron-Aware Data Selection In Instruction Tuning For Large Language Models

2026-03-13

AI Summary

This work addresses the challenge of efficiently selecting a small, high-quality subset from massive instruction-tuning data to enhance either specific or general capabilities of large language models. The authors propose NAIT, a novel framework that leverages neuron activation patterns to guide data selection. By constructing reusable activation-based features, NAIT measures the similarity between individual samples and target capabilities without relying on external models, enabling effective identification of optimal subsets. The approach reveals the transferability of activation features across diverse capabilities and pinpoints core data samples with strong generalization potential. Empirically, using only 10% of the Alpaca-GPT4 dataset, NAIT consistently outperforms baseline methods based on external models or uncertainty-based features across multiple tasks.

Technology Category

• Machine Learning: Active Learning
• Natural Language Processing: (Large) Language Models
• Search and Optimization: Learning to Search

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Instruction Tuning (IT) has been proven to be an effective approach to unlock the powerful capabilities of large language models (LLMs). Recent studies indicate that excessive IT data can degrade LLMs performance, while carefully selecting a small subset of high-quality IT data can significantly enhance their capabilities. Therefore, identifying the most efficient subset data from the IT dataset to effectively develop either specific or general abilities in LLMs has become a critical challenge. To address this, we propose a novel and efficient framework called NAIT. NAIT evaluates the impact of IT data on LLMs performance by analyzing the similarity of neuron activation patterns between the IT dataset and the target domain capability. Specifically, NAIT captures neuron activation patterns from in-domain datasets of target domain capabilities to construct reusable and transferable neuron activation features. It then evaluates and selects optimal samples based on the similarity between candidate samples and the expected activation features of the target capabilities. Experimental results show that training on the 10\% Alpaca-GPT4 IT data subset selected by NAIT consistently outperforms methods that rely on external advanced models or uncertainty-based features across various tasks. Our findings also reveal the transferability of neuron activation features across different capabilities of LLMs. In particular, IT data with more logical reasoning and programmatic features possesses strong general transferability, enabling models to develop stronger capabilities across multiple tasks, while a stable core subset of data is sufficient to consistently activate fundamental model capabilities and universally improve performance across diverse tasks.

Problem

• Instruction Tuning
• Data Selection
• Large Language Models
• Neuron Activation
• Transferability

Innovation

• neuron activation patterns
• data selection
• instruction tuning
• transferability
• large language models