Abstract
Pre-trained language models (PLMs) achieve remarkable performance on many downstream tasks, but may fail in giving reliable estimates of their predictive uncertainty. Given the lack of a comprehensive understanding of PLMs calibration, we take a close look into this new research problem, aiming to answer two questions: (1) Do PLMs learn to become calibrated in the training process? (2) How effective are existing calibration methods? For the first question, we conduct fine-grained control experiments to study the dynamic change in PLMs' calibration performance in training. We consider six factors as control variables, including dataset difficulty, available training samples, training steps, the number of tunable parameters, model scale, and pretraining. In experiments, we observe a consistent change in calibration performance across six factors. We find that PLMs don't learn to become calibrated in training, evidenced by the continual increase in confidence, no matter the predictions are correct or not. We highlight that our finding presents some contradiction with two established conclusions: (a) Larger PLMs are more calibrated; (b) Pretraining improves model calibration. Next, we study the effectiveness of existing calibration methods in mitigating the overconfidence issue, in both in-distribution and various out-of-distribution settings. Besides unlearnable calibration methods, we adapt two recently proposed learnable methods that directly collect data to train models to have reasonable confidence estimations. Also, we propose extended learnable methods based on existing ones to further improve or maintain PLMs calibration without sacrificing the original task performance. Experimental results show that learnable methods significantly reduce PLMs' confidence in wrong predictions, and our methods exhibit superior performance compared with previous methods.
Keyword: confidence
A Close Look into the Calibration of Pre-trained Language Models
Abstract
Pre-trained language models (PLMs) achieve remarkable performance on many downstream tasks, but may fail in giving reliable estimates of their predictive uncertainty. Given the lack of a comprehensive understanding of PLMs calibration, we take a close look into this new research problem, aiming to answer two questions: (1) Do PLMs learn to become calibrated in the training process? (2) How effective are existing calibration methods? For the first question, we conduct fine-grained control experiments to study the dynamic change in PLMs' calibration performance in training. We consider six factors as control variables, including dataset difficulty, available training samples, training steps, the number of tunable parameters, model scale, and pretraining. In experiments, we observe a consistent change in calibration performance across six factors. We find that PLMs don't learn to become calibrated in training, evidenced by the continual increase in confidence, no matter the predictions are correct or not. We highlight that our finding presents some contradiction with two established conclusions: (a) Larger PLMs are more calibrated; (b) Pretraining improves model calibration. Next, we study the effectiveness of existing calibration methods in mitigating the overconfidence issue, in both in-distribution and various out-of-distribution settings. Besides unlearnable calibration methods, we adapt two recently proposed learnable methods that directly collect data to train models to have reasonable confidence estimations. Also, we propose extended learnable methods based on existing ones to further improve or maintain PLMs calibration without sacrificing the original task performance. Experimental results show that learnable methods significantly reduce PLMs' confidence in wrong predictions, and our methods exhibit superior performance compared with previous methods.
ActGraph: Prioritization of Test Cases Based on Deep Neural Network Activation Graph
Abstract
Widespread applications of deep neural networks (DNNs) benefit from DNN testing to guarantee their quality. In the DNN testing, numerous test cases are fed into the model to explore potential vulnerabilities, but they require expensive manual cost to check the label. Therefore, test case prioritization is proposed to solve the problem of labeling cost, e.g., activation-based and mutation-based prioritization methods. However, most of them suffer from limited scenarios (i.e. high confidence adversarial or false positive cases) and high time complexity. To address these challenges, we propose the concept of the activation graph from the perspective of the spatial relationship of neurons. We observe that the activation graph of cases that triggers the models' misbehavior significantly differs from that of normal cases. Motivated by it, we design a test case prioritization method based on the activation graph, ActGraph, by extracting the high-order node features of the activation graph for prioritization. ActGraph explains the difference between the test cases to solve the problem of scenario limitation. Without mutation operations, ActGraph is easy to implement, leading to lower time complexity. Extensive experiments on three datasets and four models demonstrate that ActGraph has the following key characteristics. (i) Effectiveness and generalizability: ActGraph shows competitive performance in all of the natural, adversarial and mixed scenarios, especially in RAUC-100 improvement (~1.40). (ii) Efficiency: ActGraph does not use complex mutation operations and runs in less time (~1/50) than the state-of-the-art method.
Keyword: scaling
Transfer Learning with Kernel Methods
Authors: Adityanarayanan Radhakrishnan, Max Ruiz Luyten, Neha Prasad, Caroline Uhler
Abstract
Transfer learning refers to the process of adapting a model trained on a source task to a target task. While kernel methods are conceptually and computationally simple machine learning models that are competitive on a variety of tasks, it has been unclear how to perform transfer learning for kernel methods. In this work, we propose a transfer learning framework for kernel methods by projecting and translating the source model to the target task. We demonstrate the effectiveness of our framework in applications to image classification and virtual drug screening. In particular, we show that transferring modern kernels trained on large-scale image datasets can result in substantial performance increase as compared to using the same kernel trained directly on the target task. In addition, we show that transfer-learned kernels allow a more accurate prediction of the effect of drugs on cancer cell lines. For both applications, we identify simple scaling laws that characterize the performance of transfer-learned kernels as a function of the number of target examples. We explain this phenomenon in a simplified linear setting, where we are able to derive the exact scaling laws. By providing a simple and effective transfer learning framework for kernel methods, our work enables kernel methods trained on large datasets to be easily adapted to a variety of downstream target tasks.
Modelling black-box audio effects with time-varying feature modulation
Authors: Marco Comunità, Christian J. Steinmetz, Huy Phan, Joshua D. Reiss
Abstract
Deep learning approaches for black-box modelling of audio effects have shown promise, however, the majority of existing work focuses on nonlinear effects with behaviour on relatively short time-scales, such as guitar amplifiers and distortion. While recurrent and convolutional architectures can theoretically be extended to capture behaviour at longer time scales, we show that simply scaling the width, depth, or dilation factor of existing architectures does not result in satisfactory performance when modelling audio effects such as fuzz and dynamic range compression. To address this, we propose the integration of time-varying feature-wise linear modulation into existing temporal convolutional backbones, an approach that enables learnable adaptation of the intermediate activations. We demonstrate that our approach more accurately captures long-range dependencies for a range of fuzz and compressor implementations across both time and frequency domain metrics. We provide sound examples, source code, and pretrained models to faciliate reproducibility.
Keyword: calibration
A Close Look into the Calibration of Pre-trained Language Models
Abstract
Pre-trained language models (PLMs) achieve remarkable performance on many downstream tasks, but may fail in giving reliable estimates of their predictive uncertainty. Given the lack of a comprehensive understanding of PLMs calibration, we take a close look into this new research problem, aiming to answer two questions: (1) Do PLMs learn to become calibrated in the training process? (2) How effective are existing calibration methods? For the first question, we conduct fine-grained control experiments to study the dynamic change in PLMs' calibration performance in training. We consider six factors as control variables, including dataset difficulty, available training samples, training steps, the number of tunable parameters, model scale, and pretraining. In experiments, we observe a consistent change in calibration performance across six factors. We find that PLMs don't learn to become calibrated in training, evidenced by the continual increase in confidence, no matter the predictions are correct or not. We highlight that our finding presents some contradiction with two established conclusions: (a) Larger PLMs are more calibrated; (b) Pretraining improves model calibration. Next, we study the effectiveness of existing calibration methods in mitigating the overconfidence issue, in both in-distribution and various out-of-distribution settings. Besides unlearnable calibration methods, we adapt two recently proposed learnable methods that directly collect data to train models to have reasonable confidence estimations. Also, we propose extended learnable methods based on existing ones to further improve or maintain PLMs calibration without sacrificing the original task performance. Experimental results show that learnable methods significantly reduce PLMs' confidence in wrong predictions, and our methods exhibit superior performance compared with previous methods.
Using coevolution and substitution of the fittest for health and well-being recommender systems
Authors: Hugo Alcaraz-Herrera, John Cartlidge
Subjects: Neural and Evolutionary Computing (cs.NE); Artificial Intelligence (cs.AI)
Abstract
This research explores substitution of the fittest (SF), a technique designed to counteract the problem of disengagement in two-population competitive coevolutionary genetic algorithms. SF is domain-independent and requires no calibration. We first perform a controlled comparative evaluation of SF's ability to maintain engagement and discover optimal solutions in a minimal toy domain. Experimental results demonstrate that SF is able to maintain engagement better than other techniques in the literature. We then address the more complex real-world problem of evolving recommendations for health and well-being. We introduce a coevolutionary extension of EvoRecSys, a previously published evolutionary recommender system. We demonstrate that SF is able to maintain engagement better than other techniques in the literature, and the resultant recommendations using SF are higher quality and more diverse than those produced by EvoRecSys.
Keyword: out of distribution detection
There is no result
Keyword: out-of-distribution detection
There is no result
Keyword: expected calibration error
There is no result
Keyword: overconfident
There is no result
Keyword: overconfidence
A Close Look into the Calibration of Pre-trained Language Models
Keyword: confidence
A Close Look into the Calibration of Pre-trained Language Models
ActGraph: Prioritization of Test Cases Based on Deep Neural Network Activation Graph
Keyword: scaling
Transfer Learning with Kernel Methods
Modelling black-box audio effects with time-varying feature modulation
Keyword: calibration
A Close Look into the Calibration of Pre-trained Language Models
Using coevolution and substitution of the fittest for health and well-being recommender systems