NoisyICL: A Little Noise in Model Parameters Calibrates In-context Learning (2402.05515v2)

Abstract: In-Context Learning (ICL) is suffering from unsatisfactory performance and under-calibration due to high prior bias and unfaithful confidence. Some previous works fine-tuned LLMs for better ICL performance with enormous datasets and computing costs. In this paper, we propose NoisyICL, simply perturbing the model parameters by random noises to strive for better performance and calibration. Our experiments on two models and 12 downstream datasets show that NoisyICL can help ICL produce more accurate predictions. Our further analysis indicates that NoisyICL enables the model to provide more fair predictions, and also with more faithful confidence. Therefore, we believe that NoisyICL is an effective calibration of ICL. Our experimental code is uploaded to Github.