Learning to Encode Position for Transformer with Continuous Dynamical Model

We introduce a new way of learning to encode position information for non-recurrent models, such as Transformer models. Unlike RNN and LSTM, which contain inductive bias by loading the input tokens sequentially, non-recurrent models are less sensitive to position. The main reason is that position information among input units is not inherently encoded, i.e., the models are permutation equivalent; this problem justifies why all of the existing models are accompanied by a sinusoidal encoding/embedding layer at the input. However, this solution has clear limitations: the sinusoidal encoding is not flexible enough as it is manually designed and does not contain any learnable parameters, whereas the position embedding restricts the maximum length of input sequences. It is thus desirable to design a new position layer that contains learnable parameters to adjust to different datasets and different architectures. At the same time, we would also like the encodings to extrapolate in accordance with the variable length of inputs. In our proposed solution, we borrow from the recent Neural ODE approach, which may be viewed as a versatile continuous version of a ResNet. This model is capable of modeling many kinds of dynamical systems. We model the evolution of encoded results along position index by such a dynamical system, thereby overcoming the above limitations of existing methods. We evaluate our new position layers on a variety of neural machine translation and language understanding tasks, the experimental results show consistent improvements over the baselines.

Automatic Perturbation Analysis on General Computational Graphs

Linear relaxation based perturbation analysis for neural networks, which aims to compute tight linear bounds of output neurons given a certain amount of input perturbation, has become a core component in robustness verification and certified defense. However, the majority of linear relaxation based methods only consider feed-forward ReLU networks. While several works extended them to relatively complicated networks, they often need tedious manual derivations and implementation which are arduous and error-prone. Their limited flexibility makes it difficult to handle more complicated tasks. In this paper, we take a significant leap by developing an automatic perturbation analysis algorithm to enable perturbation analysis on any neural network structure, and its computation can be done automatically in a similar manner as the back-propagation algorithm for gradient computation. The main idea is to express a network as a computational graph and then generalize linear relaxation algorithms such as CROWN as a graph algorithm. Our algorithm itself is differentiable and integrated with PyTorch, which allows to optimize network parameters to reshape bounds into desired specifications, enabling automatic robustness verification and certified defense. In particular, we demonstrate a few tasks that are not easily achievable without an automatic framework. We first perform certified robust training and robustness verification for complex natural language models which could be challenging with manual derivation and implementation. We further show that our algorithm can be used for tasks beyond certified defense - we create a neural network with a provably flat optimization landscape and study its generalization capability, and we show that this network can preserve accuracy better after aggressive weight quantization. Code is available at https://github.com/KaidiXu/auto_LiRPA.

CAT: Customized Adversarial Training for Improved Robustness

Adversarial training has become one of the most effective methods for improving robustness of neural networks. However, it often suffers from poor generalization on both clean and perturbed data. In this paper, we propose a new algorithm, named Customized Adversarial Training (CAT), which adaptively customizes the perturbation level and the corresponding label for each training sample in adversarial training. We show that the proposed algorithm achieves better clean and robust accuracy than previous adversarial training methods through extensive experiments.

Robustness Verification for Transformers

Robustness verification that aims to formally certify the prediction behavior of neural networks has become an important tool for understanding model behavior and obtaining safety guarantees. However, previous methods can usually only handle neural networks with relatively simple architectures. In this paper, we consider the robustness verification problem for Transformers. Transformers have complex self-attention layers that pose many challenges for verification, including cross-nonlinearity and cross-position dependency, which have not been discussed in previous works. We resolve these challenges and develop the first robustness verification algorithm for Transformers. The certified robustness bounds computed by our method are significantly tighter than those by naive Interval Bound Propagation. These bounds also shed light on interpreting Transformers as they consistently reflect the importance of different words in sentiment analysis.

MACER: Attack-free and Scalable Robust Training via Maximizing Certified Radius

Adversarial training is one of the most popular ways to learn robust models but is usually attack-dependent and time costly. In this paper, we propose the MACER algorithm, which learns robust models without using adversarial training but performs better than all existing provable l2-defenses. Recent work shows that randomized smoothing can be used to provide a certified l2 radius to smoothed classifiers, and our algorithm trains provably robust smoothed classifiers via MAximizing the CErtified Radius (MACER). The attack-free characteristic makes MACER faster to train and easier to optimize. In our experiments, we show that our method can be applied to modern deep neural networks on a wide range of datasets, including Cifar-10, ImageNet, MNIST, and SVHN. For all tasks, MACER spends less training time than state-of-the-art adversarial training algorithms, and the learned models achieve larger average certified radius.

Multiscale Non-stationary Stochastic Bandits

Classic contextual bandit algorithms for linear models, such as LinUCB, assume that the reward distribution for an arm is modeled by a stationary linear regression. When the linear regression model is non-stationary over time, the regret of LinUCB can scale linearly with time. In this paper, we propose a novel multiscale changepoint detection method for the non-stationary linear bandit problems, called Multiscale-LinUCB, which actively adapts to the changing environment. We also provide theoretical analysis of regret bound for Multiscale-LinUCB algorithm. Experimental results show that our proposed Multiscale-LinUCB algorithm outperforms other state-of-the-art algorithms in non-stationary contextual environments.

Stabilizing Differentiable Architecture Search via Perturbation-based Regularization

Differentiable architecture search (DARTS) is a prevailing NAS solution to identify architectures. Based on the continuous relaxation of the architecture space, DARTS learns a differentiable architecture weight and largely reduces the search cost. However, its stability and generalizability have been challenged for yielding deteriorating architectures as the search proceeds. We find that the precipitous validation loss landscape, which leads to a dramatic performance drop when distilling the final architecture, is an essential factor that causes instability. Based on this observation, we propose a perturbation-based regularization, named SmoothDARTS (SDARTS), to smooth the loss landscape and improve the generalizability of DARTS. In particular, our new formulations stabilize DARTS by either random smoothing or adversarial attack. The search trajectory on NAS-Bench-1Shot1 demonstrates the effectiveness of our approach and due to the improved stability, we achieve performance gain across various search spaces on 4 datasets. Furthermore, we mathematically show that SDARTS implicitly regularizes the Hessian norm of the validation loss, which accounts for a smoother loss landscape and improved performance. The code is available at https://github.com/xiangning-chen/SmoothDARTS.

GraphDefense: Towards Robust Graph Convolutional Networks

In this paper, we study the robustness of graph convolutional networks (GCNs). Despite the good performance of GCNs on graph semi-supervised learning tasks, previous works have shown that the original GCNs are very unstable to adversarial perturbations. In particular, we can observe a severe performance degradation by slightly changing the graph adjacency matrix or the features of a few nodes, making it unsuitable for security-critical applications. Inspired by the previous works on adversarial defense for deep neural networks, and especially adversarial training algorithm, we propose a method called GraphDefense to defend against the adversarial perturbations. In addition, for our defense method, we could still maintain semi-supervised learning settings, without a large label rate. We also show that adversarial training in features is equivalent to adversarial training for edges with a small perturbation. Our experiments show that the proposed defense methods successfully increase the robustness of Graph Convolutional Networks. Furthermore, we show that with careful design, our proposed algorithm can scale to large graphs, such as Reddit dataset.

Enhancing Certifiable Robustness via a Deep Model Ensemble

We propose an algorithm to enhance certified robustness of a deep model ensemble by optimally weighting each base model. Unlike previous works on using ensembles to empirically improve robustness, our algorithm is based on optimizing a guaranteed robustness certificate of neural networks. Our proposed ensemble framework with certified robustness, RobBoost, formulates the optimal model selection and weighting task as an optimization problem on a lower bound of classification margin, which can be efficiently solved using coordinate descent. Experiments show that our algorithm can form a more robust ensemble than naively averaging all available models using robustly trained MNIST or CIFAR base models. Additionally, our ensemble typically has better accuracy on clean (unperturbed) data. RobBoost allows us to further improve certified robustness and clean accuracy by creating an ensemble of already certified models.

A Unified Framework for Data Poisoning Attack to Graph-based Semi-supervised Learning

In this paper, we proposed a general framework for data poisoning attacks to graph-based semi-supervised learning (G-SSL). In this framework, we first unify different tasks, goals, and constraints into a single formula for data poisoning attack in G-SSL, then we propose two specialized algorithms to efficiently solve two important cases --- poisoning regression tasks under $\ell_2$-norm constraint and classification tasks under $\ell_0$-norm constraint. In the former case, we transform it into a non-convex trust region problem and show that our gradient-based algorithm with delicate initialization and update scheme finds the (globally) optimal perturbation. For the latter case, although it is an NP-hard integer programming problem, we propose a probabilistic solver that works much better than the classical greedy method. Lastly, we test our framework on real datasets and evaluate the robustness of G-SSL algorithms. For instance, on the MNIST binary classification problem (50000 training data with 50 labeled), flipping two labeled data is enough to make the model perform like random guess (around 50\% error).