Structured Gradient Guidance for Few-Shot Adaptation in Large Language Models

TL;DR

This paper introduces a gradient-guided fine-tuning approach for large language models that improves task adaptability, stability, and cross-task generalization in few-shot settings by using gradient regularization and alignment mechanisms.

Contribution

The paper proposes a novel gradient-informed fine-tuning method with regularization and alignment techniques to enhance stability and generalization in low-resource scenarios.

Findings

Outperforms existing fine-tuning strategies in accuracy and stability

Enhances cross-task generalization through gradient alignment

Demonstrates robustness across various NLP tasks and data sizes

Abstract

This paper presents a gradient-informed fine-tuning method for large language models under few-shot conditions. The goal is to enhance task adaptability and training stability when data is limited. The method builds on a base loss function and introduces two gradient-related regularization terms. The first enforces gradient direction consistency to guide parameter updates along task-relevant directions and prevent drift. The second controls gradient magnitude to avoid abnormal updates. Together, these components support a more efficient and stable optimization path. To further improve cross-task generalization, the method incorporates a gradient alignment mechanism. This mechanism measures the consistency between optimization directions of the source and target tasks. It enhances fine-tuning performance in multi-task and cross-domain scenarios. Across various natural language understanding tasks, the method outperforms existing fine-tuning strategies in average accuracy, gradient stability, and directional alignment. Empirical evaluations under different sample sizes and domain-specific tasks confirm the method's robustness and broad applicability in low-resource environments. In particular, the method shows clear advantages in controlling parameter update paths. The results demonstrate that a gradient-based fine-tuning framework can effectively leverage the representational power of large language models. It ensures training stability while reducing dependence on large volumes of labeled data.