Weight-sharing supernet has become a vital component for performance estimation in the state-of-the-art (SOTA) neural architecture search (NAS) frameworks. Although supernet can directly generate different subnetworks without retraining, there is no guarantee for the quality of these subnetworks because of weight sharing. In NLP tasks such as machine translation and pre-trained language modeling, we observe that given the same model architecture, there is a large performance gap between supernet and training from scratch. Hence, supernet cannot be directly used and retraining is necessary after finding the optimal architectures. In this work, we propose mixture-of-supernets, a generalized supernet formulation where mixture-of-experts (MoE) is adopted to enhance the expressive power of the supernet model, with negligible training overhead. In this way, different subnetworks do not share the model weights directly, but through an architecture-based routing mechanism. As a result, model weights of different subnetworks are customized towards their specific architectures and the weight generation is learned by gradient descent. Compared to existing weight-sharing supernet for NLP, our method can minimize the retraining time, greatly improving training efficiency. In addition, the proposed method achieves the SOTA performance in NAS for building fast machine translation models, yielding better latency-BLEU tradeoff compared to HAT, state-of-the-art NAS for MT. We also achieve the SOTA performance in NAS for building memory-efficient task-agnostic BERT models, outperforming NAS-BERT and AutoDistil in various model sizes.
The generalization of neural networks is a central challenge in machine learning, especially concerning the performance under distributions that differ from training ones. Current methods, mainly based on the data-driven paradigm such as data augmentation, adversarial training, and noise injection, may encounter limited generalization due to model non-smoothness. In this paper, we propose to investigate generalization from a Partial Differential Equation (PDE) perspective, aiming to enhance it directly through the underlying function of neural networks, rather than focusing on adjusting input data. Specifically, we first establish the connection between neural network generalization and the smoothness of the solution to a specific PDE, namely ``transport equation''. Building upon this, we propose a general framework that introduces adaptive distributional diffusion into transport equation to enhance the smoothness of its solution, thereby improving generalization. In the context of neural networks, we put this theoretical framework into practice as PDE+ (\textbf{PDE} with \textbf{A}daptive \textbf{D}istributional \textbf{D}iffusion) which diffuses each sample into a distribution covering semantically similar inputs. This enables better coverage of potentially unobserved distributions in training, thus improving generalization beyond merely data-driven methods. The effectiveness of PDE+ is validated in extensive settings, including clean samples and various corruptions, demonstrating its superior performance compared to SOTA methods.
The development of knowledge graph (KG) applications has led to a rising need for entity alignment (EA) between heterogeneous KGs that are extracted from various sources. Recently, graph neural networks (GNNs) have been widely adopted in EA tasks due to GNNs' impressive ability to capture structure information. However, we have observed that the oversimplified settings of the existing common EA datasets are distant from real-world scenarios, which obstructs a full understanding of the advancements achieved by recent methods. This phenomenon makes us ponder: Do existing GNN-based EA methods really make great progress? In this paper, to study the performance of EA methods in realistic settings, we focus on the alignment of highly heterogeneous KGs (HHKGs) (e.g., event KGs and general KGs) which are different with regard to the scale and structure, and share fewer overlapping entities. First, we sweep the unreasonable settings, and propose two new HHKG datasets that closely mimic real-world EA scenarios. Then, based on the proposed datasets, we conduct extensive experiments to evaluate previous representative EA methods, and reveal interesting findings about the progress of GNN-based EA methods. We find that the structural information becomes difficult to exploit but still valuable in aligning HHKGs. This phenomenon leads to inferior performance of existing EA methods, especially GNN-based methods. Our findings shed light on the potential problems resulting from an impulsive application of GNN-based methods as a panacea for all EA datasets. Finally, we introduce a simple but effective method: Simple-HHEA, which comprehensively utilizes entity name, structure, and temporal information. Experiment results show Simple-HHEA outperforms previous models on HHKG datasets.
Machine unlearning aims to erase the impact of specific training samples upon deleted requests from a trained model. Re-training the model on the retained data after deletion is an effective but not efficient way due to the huge number of model parameters and re-training samples. To speed up, a natural way is to reduce such parameters and samples. However, such a strategy typically leads to a loss in model performance, which poses the challenge that increasing the unlearning efficiency while maintaining acceptable performance. In this paper, we present a novel network, namely \textit{LegoNet}, which adopts the framework of ``fixed encoder + multiple adapters''. We fix the encoder~(\ie the backbone for representation learning) of LegoNet to reduce the parameters that need to be re-trained during unlearning. Since the encoder occupies a major part of the model parameters, the unlearning efficiency is significantly improved. However, fixing the encoder empirically leads to a significant performance drop. To compensate for the performance loss, we adopt the ensemble of multiple adapters, which are independent sub-models adopted to infer the prediction by the encoding~(\ie the output of the encoder). Furthermore, we design an activation mechanism for the adapters to further trade off unlearning efficiency against model performance. This mechanism guarantees that each sample can only impact very few adapters, so during unlearning, parameters and samples that need to be re-trained are both reduced. The empirical experiments verify that LegoNet accomplishes fast and exact unlearning while maintaining acceptable performance, synthetically outperforming unlearning baselines.
Self-attention based transformer models have been dominating many computer vision tasks in the past few years. Their superb model qualities heavily depend on the excessively large labeled image datasets. In order to reduce the reliance on large labeled datasets, reconstruction based masked autoencoders are gaining popularity, which learn high quality transferable representations from unlabeled images. For the same purpose, recent weakly supervised image pretraining methods explore language supervision from text captions accompanying the images. In this work, we propose masked image pretraining on language assisted representation, dubbed as MILAN. Instead of predicting raw pixels or low level features, our pretraining objective is to reconstruct the image features with substantial semantic signals that are obtained using caption supervision. Moreover, to accommodate our reconstruction target, we propose a more efficient prompting decoder architecture and a semantic aware mask sampling mechanism, which further advance the transfer performance of the pretrained model. Experimental results demonstrate that MILAN delivers higher accuracy than the previous works. When the masked autoencoder is pretrained and finetuned on ImageNet-1K dataset with an input resolution of 224x224, MILAN achieves a top-1 accuracy of 85.4% on ViTB/16, surpassing previous state-of-the-arts by 1%. In the downstream semantic segmentation task, MILAN achieves 52.7 mIoU using ViT-B/16 backbone on ADE20K dataset, outperforming previous masked pretraining results by 4 points.
Learned recommender systems may inadvertently leak information about their training data, leading to privacy violations. We investigate privacy threats faced by recommender systems through the lens of membership inference. In such attacks, an adversary aims to infer whether a user's data is used to train the target recommender. To achieve this, previous work has used a shadow recommender to derive training data for the attack model, and then predicts the membership by calculating difference vectors between users' historical interactions and recommended items. State-of-the-art methods face two challenging problems: (1) training data for the attack model is biased due to the gap between shadow and target recommenders, and (2) hidden states in recommenders are not observational, resulting in inaccurate estimations of difference vectors. To address the above limitations, we propose a Debiasing Learning for Membership Inference Attacks against recommender systems (DL-MIA) framework that has four main components: (1) a difference vector generator, (2) a disentangled encoder, (3) a weight estimator, and (4) an attack model. To mitigate the gap between recommenders, a variational auto-encoder (VAE) based disentangled encoder is devised to identify recommender invariant and specific features. To reduce the estimation bias, we design a weight estimator, assigning a truth-level score for each difference vector to indicate estimation accuracy. We evaluate DL-MIA against both general recommenders and sequential recommenders on three real-world datasets. Experimental results show that DL-MIA effectively alleviates training and estimation biases simultaneously, and achieves state-of-the-art attack performance.
Recently, privacy issues in web services that rely on users' personal data have raised great attention. Unlike existing privacy-preserving technologies such as federated learning and differential privacy, we explore another way to mitigate users' privacy concerns, giving them control over their own data. For this goal, we propose a privacy aware recommendation framework that gives users delicate control over their personal data, including implicit behaviors, e.g., clicks and watches. In this new framework, users can proactively control which data to disclose based on the trade-off between anticipated privacy risks and potential utilities. Then we study users' privacy decision making under different data disclosure mechanisms and recommendation models, and how their data disclosure decisions affect the recommender system's performance. To avoid the high cost of real-world experiments, we apply simulations to study the effects of our proposed framework. Specifically, we propose a reinforcement learning algorithm to simulate users' decisions (with various sensitivities) under three proposed platform mechanisms on two datasets with three representative recommendation models. The simulation results show that the platform mechanisms with finer split granularity and more unrestrained disclosure strategy can bring better results for both end users and platforms than the "all or nothing" binary mechanism adopted by most real-world applications. It also shows that our proposed framework can effectively protect users' privacy since they can obtain comparable or even better results with much less disclosed data.
Channel pruning has been broadly recognized as an effective technique to reduce the computation and memory cost of deep convolutional neural networks. However, conventional pruning methods have limitations in that: they are restricted to pruning process only, and they require a fully pre-trained large model. Such limitations may lead to sub-optimal model quality as well as excessive memory and training cost. In this paper, we propose a novel Channel Exploration methodology, dubbed as CHEX, to rectify these problems. As opposed to pruning-only strategy, we propose to repeatedly prune and regrow the channels throughout the training process, which reduces the risk of pruning important channels prematurely. More exactly: From intra-layer's aspect, we tackle the channel pruning problem via a well known column subset selection (CSS) formulation. From inter-layer's aspect, our regrowing stages open a path for dynamically re-allocating the number of channels across all the layers under a global channel sparsity constraint. In addition, all the exploration process is done in a single training from scratch without the need of a pre-trained large model. Experimental results demonstrate that CHEX can effectively reduce the FLOPs of diverse CNN architectures on a variety of computer vision tasks, including image classification, object detection, instance segmentation, and 3D vision. For example, our compressed ResNet-50 model on ImageNet dataset achieves 76% top1 accuracy with only 25% FLOPs of the original ResNet-50 model, outperforming previous state-of-the-art channel pruning methods. The checkpoints and code are available at here .
As the final stage of the multi-stage recommender system (MRS), re-ranking directly affects user experience and satisfaction by rearranging the input ranking lists, and thereby plays a critical role in MRS. With the advances in deep learning, neural re-ranking has become a trending topic and been widely applied in industrial applications. This review aims at integrating re-ranking algorithms into a broader picture, and paving ways for more comprehensive solutions for future research. For this purpose, we first present a taxonomy of current methods on neural re-ranking. Then we give a description of these methods along with the historic development according to their objectives. The network structure, personalization, and complexity are also discussed and compared. Next, we provide benchmarks of the major neural re-ranking models and quantitatively analyze their re-ranking performance. Finally, the review concludes with a discussion on future prospects of this field. A list of papers discussed in this review, the benchmark datasets, our re-ranking library LibRerank, and detailed parameter settings are publicly available at https://github.com/LibRerank-Community/LibRerank.
Recommender systems provide essential web services by learning users' personal preferences from collected data. However, in many cases, systems also need to forget some training data. From the perspective of privacy, several privacy regulations have recently been proposed, requiring systems to eliminate any impact of the data whose owner requests to forget. From the perspective of utility, if a system's utility is damaged by some bad data, the system needs to forget these data to regain utility. From the perspective of usability, users can delete noise and incorrect entries so that a system can provide more useful recommendations. While unlearning is very important, it has not been well-considered in existing recommender systems. Although there are some researches have studied the problem of machine unlearning in the domains of image and text data, existing methods can not been directly applied to recommendation as they are unable to consider the collaborative information. In this paper, we propose RecEraser, a general and efficient machine unlearning framework tailored to recommendation task. The main idea of RecEraser is to partition the training set into multiple shards and train a constituent model for each shard. Specifically, to keep the collaborative information of the data, we first design three novel data partition algorithms to divide training data into balanced groups based on their similarity. Then, considering that different shard models do not uniformly contribute to the final prediction, we further propose an adaptive aggregation method to improve the global model utility. Experimental results on three public benchmarks show that RecEraser can not only achieve efficient unlearning, but also outperform the state-of-the-art unlearning methods in terms of model utility. The source code can be found at https://github.com/chenchongthu/Recommendation-Unlearning