Towards Alignment-Centric Paradigm: A Survey of Instruction Tuning in Large Language Models

TL;DR

This survey reviews instruction tuning techniques for large language models, covering data collection, fine-tuning methods, evaluation challenges, and future directions to improve alignment with human goals.

Contribution

It provides a comprehensive categorization of data paradigms, fine-tuning strategies, and evaluation protocols, highlighting recent advances and future research directions in instruction tuning.

Findings

Data construction paradigms vary in quality and scalability.

Lightweight fine-tuning methods like LoRA improve efficiency.

Evaluation of faithfulness and safety remains challenging.

Abstract

Instruction tuning is a pivotal technique for aligning large language models (LLMs) with human intentions, safety constraints, and domain-specific requirements. This survey provides a comprehensive overview of the full pipeline, encompassing (i) data collection methodologies, (ii) full-parameter and parameter-efficient fine-tuning strategies, and (iii) evaluation protocols. We categorized data construction into three major paradigms: expert annotation, distillation from larger models, and self-improvement mechanisms, each offering distinct trade-offs between quality, scalability, and resource cost. Fine-tuning techniques range from conventional supervised training to lightweight approaches, such as low-rank adaptation (LoRA) and prefix tuning, with a focus on computational efficiency and model reusability. We further examine the challenges of evaluating faithfulness, utility, and safety across multilingual and multimodal scenarios, highlighting the emergence of domain-specific benchmarks in healthcare, legal, and financial applications. Finally, we discuss promising directions for automated data generation, adaptive optimization, and robust evaluation frameworks, arguing that a closer integration of data, algorithms, and human feedback is essential for advancing instruction-tuned LLMs. This survey aims to serve as a practical reference for researchers and practitioners seeking to design LLMs that are both effective and reliably aligned with human intentions.