# Massive Supervised Fine-tuning Experiments Reveal How Data, Layer, and Training Factors Shape LLM Alignment Quality

**Authors:** Yuto Harada, Yusuke Yamauchi, Yusuke Oda, Yohei Oseki, Yusuke Miyao, Yu Takagi  
**arXiv:** [2506.14681](https://arxiv.org/abs/2506.14681)  
**Date:** 2025-06-17  
**Summary page:** https://www.summarizepaper.com/fr/paper/2506.14681v1/

## Abstract

Supervised fine-tuning (SFT) is a critical step in aligning large language models (LLMs) with human instructions and values, yet many aspects of SFT remain poorly understood. We trained a wide range of base models on a variety of datasets including code generation, mathematical reasoning, and general-domain tasks, resulting in 1,000+ SFT models under controlled conditions. We then identified the dataset properties that matter most and examined the layer-wise modifications introduced by SFT. Our findings reveal that some training-task synergies persist across all models while others vary substantially, emphasizing the importance of model-specific strategies. Moreover, we demonstrate that perplexity consistently predicts SFT effectiveness--often surpassing superficial similarity between trained data and benchmark--and that mid-layer weight changes correlate most strongly with performance gains. We will release these 1,000+ SFT models and benchmark results to accelerate further research.