Harnessing non-adversarial robustness in large language models
The work presents an approach for addressing the challenge of robustness in Large Language Models (LLMs) to alterations and potential errors caused by semantically similar but textually different prompts. Recent works have shown that these kinds of prompt variations can significantly impact the performance of LLMs on tasks. The central question is: can LLMs' robustness to semantically-neutral prompt alterations be acquired without expensive retraining of the entire model? We address this question both theoretically and through experiments. Our theoretical analysis reveals a crucial factor impacting model robustness - a systematic expected shift or perturbation-induced bias in neural network module outputs. Motivated by this analysis, we show that robustness can be achieved via a simple fine-tuning process: debiasing for robustness. We identify conditions when debiasing helps and when it does not, and demonstrate, through both theory and extensive experiments, that debiasing for robustness may indeed be a quick and efficient tool to enhance robustness and provide certification against random prompt perturbations.
Code (0)
등록된 구현이 없습니다.
Tasks
Adversarial RobustnessSimilar Papers 제목 키워드 기반
Adapters Mixup: Mixing Parameter-Efficient Adapters to Enhance the Adversarial Robustness of Fine-tuned Pre-trained Text Classifiers
Existing works show that augmenting the training data of pre-trained language models (PLMs) for classification tasks fine-tuned via parameter-efficient fine-tuning methods (PEFT) using both clean and adversarial examples…
Adversarial Robustnessparameter-efficient fine-tuningtext-classificationText ClassificationHarnessing the Power of Adversarial Prompting and Large Language Models for Robust Hypothesis Generation in Astronomy
This study investigates the application of Large Language Models (LLMs), specifically GPT-4, within Astronomy. We employ in-context prompting, supplying the model with up to 1000 papers from the NASA Astrophysics Data Sy…
AstronomyStable and Efficient Adversarial Training through Local Linearization
There has been a recent surge in single-step adversarial training as it shows robustness and efficiency. However, a phenomenon referred to as ``catastrophic overfitting" has been observed, which is prevalent in single-st…
Harnessing Perceptual Adversarial Patches for Crowd Counting
Crowd counting, which has been widely adopted for estimating the number of people in safety-critical scenes, is shown to be vulnerable to adversarial examples in the physical world (e.g., adversarial patches). Though har…
Crowd CountingHarnessing the Vulnerability of Latent Layers in Adversarially Trained Models
Neural networks are vulnerable to adversarial attacks -- small visually imperceptible crafted noise which when added to the input drastically changes the output. The most effective method of defending against these adver…
Adversarial Attack