Small Data, Big Noise: Adversarial Training for Robust Parameter-Efficient Fine-Tuning

2026-06-09

AI Summary

This work addresses the limited generalization of parameter-efficient fine-tuning (PEFT) under few-shot and high-noise conditions. We present SDBN, a unified framework that integrates adversarial training with PEFT to enhance robustness without increasing model parameters. SDBN introduces two novel variants based on discrete uncertainty sets: SDBN-h for character-level perturbations and SDBN-p for linguistic variations in generative tasks. The framework further incorporates gradient-guided worst-case example selection and leverages perturbations generated by large language models for efficient optimization. Extensive experiments demonstrate that SDBN consistently outperforms existing methods across multiple benchmarks, achieving particularly strong performance in low-resource settings and under word- and character-level noise, with only a marginal increase in computational overhead.

Abstract

Parameter-Efficient Fine-Tuning (PEFT) has become essential for adapting foundation models to downstream NLP tasks. However, current PEFT methods often struggle with robustness to noise and performance degradation on limited training data. We propose SDBN (Small Data Big Noise), a unified framework that brings adversarial training to PEFT - a combination that remains less studied in the PEFT setting despite its complementary strengths - to enhance model robustness and generalization, outperforming alternative approaches. We also introduce two variants of the method that use discrete uncertainty sets: SDBN-h, which enumerates character-level edits and selects worst-case variants using gradients, and SDBN-p, which uses LLM-generated variants for robust optimization in generative tasks. Experiments across multiple benchmarks reveal substantial improvements, particularly in low-resource settings and under both word-level and character-level corruptions. This framework addresses the less explored intersection of adversarial training and parameter-efficient adaptation, without introducing additional parameters or only modest computational overhead, making PEFT deployments more reliable in real-world scenarios where data scarcity and linguistic variability often coexist

Problem

• Parameter-Efficient Fine-Tuning
• robustness
• small data
• noise
• adversarial training

Innovation

• Adversarial Training
• Parameter-Efficient Fine-Tuning
• Robustness
• Small Data
• Discrete Uncertainty Sets