NeurIPS 2023 Papers

In this study, we propose a novel deep spatio-temporal point process model, Deep Kernel Mixture Point Processes (DKMPP), that incorporates multimodal covariate information. DKMPP is an enhanced version of Deep Mixture Point Processes (DMPP), which uses a more flexible deep kernel to model complex relationships between events and covariate data, improving the model's expressiveness. To address the intractable training procedure of DKMPP due to the non-integrable deep kernel, we utilize an integration-free method based on score matching, and further improve efficiency by adopting a scalable denoising score matching method. Our experiments demonstrate that DKMPP and its corresponding score-based estimators outperform baseline models, showcasing the advantages of incorporating covariate information, utilizing a deep kernel, and employing score-based estimators.

Existing deep learning-based computer vision methods usually operate in the spatial and frequency domains, which are two orthogonal \textbf{individual} perspectives for image processing. In this paper, we introduce a new spatial-frequency analysis tool, Fractional Fourier Transform (FRFT), to provide comprehensive \textbf{unified} spatial-frequency perspectives. The FRFT is a unified continuous spatial-frequency transform that simultaneously reflects an image's spatial and frequency representations, making it optimal for processing non-stationary image signals. We explore the properties of the FRFT for image processing and present a fast implementation of the 2D FRFT, which facilitates its widespread use. Based on these explorations, we introduce a simple yet effective operator, Multi-order FRactional Fourier Convolution (MFRFC), which exhibits the remarkable merits of processing images from more perspectives in the spatial-frequency plane. Our proposed MFRFC is a general and basic operator that can be easily integrated into various tasks for performance improvement. We experimentally evaluate the MFRFC on various computer vision tasks, including object detection, image classification, guided super-resolution, denoising, dehazing, deraining, and low-light enhancement. Our proposed MFRFC consistently outperforms baseline methods by significant margins across all tasks.

In modern deep learning, algorithmic choices (such as width, depth, and learning rate) are known to modulate nuanced resource tradeoffs. This work investigates how these complexities necessarily arise for feature learning in the presence of computational-statistical gaps. We begin by considering offline sparse parity learning, a supervised classification problem which admits a statistical query lower bound for gradient-based training of a multilayer perceptron. This lower bound can be interpreted as a *multi-resource tradeoff frontier*: successful learning can only occur if one is sufficiently rich (large model), knowledgeable (large dataset), patient (many training iterations), or lucky (many random guesses). We show, theoretically and experimentally, that sparse initialization and increasing network width yield significant improvements in sample efficiency in this setting. Here, width plays the role of parallel search: it amplifies the probability of finding "lottery ticket" neurons, which learn sparse features more sample-efficiently. Finally, we show that the synthetic sparse parity task can be useful as a proxy for real problems requiring axis-aligned feature learning. We demonstrate improved sample efficiency on tabular classification benchmarks by using wide, sparsely-initialized MLP models; these networks sometimes outperform tuned random forests.

Split Inference (SI) is an emerging deep learning paradigm that addresses computational constraints on edge devices and preserves data privacy through collaborative edge-cloud approaches. However, SI is vulnerable to Data Reconstruction Attacks (DRA), which aim to reconstruct users' private prediction instances. Existing attack methods suffer from various limitations. Optimization-based DRAs do not leverage public data effectively, while Learning-based DRAs depend heavily on auxiliary data quantity and distribution similarity. Consequently, these approaches yield unsatisfactory attack results and are sensitive to defense mechanisms. To overcome these challenges, we propose a GAN-based LAtent Space Search attack (GLASS) that harnesses abundant prior knowledge from public data using advanced StyleGAN technologies. Additionally, we introduce GLASS++ to enhance reconstruction stability. Our approach represents the first GAN-based DRA against SI, and extensive evaluation across different split points and adversary setups demonstrates its state-of-the-art performance. Moreover, we thoroughly examine seven defense mechanisms, highlighting our method's capability to reveal private information even in the presence of these defenses.

Stabilizing proteins is a foundational step in protein engineering. However, the evolutionary pressure of all extant proteins makes identifying the scarce number of mutations that will improve thermodynamic stability challenging. Deep learning has recently emerged as a powerful tool for identifying promising mutations. Existing approaches, however, are computationally expensive, as the number of model inferences scales with the number of mutations queried. Our main contribution is a simple, parallel decoding algorithm. Mutate Everything is capable of predicting the effect of all single and double mutations in one forward pass. It is even versatile enough to predict higher-order mutations with minimal computational overhead. We build Mutate Everything on top of ESM2 and AlphaFold, neither of which were trained to predict thermodynamic stability. We trained on the Mega-Scale cDNA proteolysis dataset and achieved state-of-the-art performance on single and higher-order mutations on S669, ProTherm, and ProteinGym datasets. Our code is available at https://github.com/jozhang97/MutateEverything.

Mixup is an efficient data augmentation approach that improves the generalization of neural networks by smoothing the decision boundary with mixed data. Recently, dynamic mixup methods have improved previous \textit{static} policies effectively (e.g., linear interpolation) by maximizing target-related salient regions in mixed samples, but excessive additional time costs are not acceptable. These additional computational overheads mainly come from optimizing the mixed samples according to the mixed labels. However, we found that the extra optimizing step may be redundant because label-mismatched mixed samples are informative hard mixed samples for deep models to localize discriminative features. In this paper, we thus are not trying to propose a more complicated dynamic mixup policy but rather an efficient mixup objective function with decoupled regularizer, named decoupled mixup (DM). The primary effect is that DM can adaptively utilize those hard mixed samples to mine discriminative features without losing the original smoothness of mixup. As a result, DM enables static mixup methods to achieve comparable or even exceed the performance of dynamic methods without any extra computation. This also leads to an interesting objective design problem for mixup training that we need to focus on both smoothing the decision boundaries and identifying discriminative features. Extensive experiments on supervised and semi-supervised learning benchmarks across seven datasets validate the effectiveness of DM.

Distributed training of Deep Learning models has been critical to many recent successes in the field. Current standard methods primarily rely on synchronous centralized algorithms which induce major communication bottlenecks and synchronization locks at scale. Decentralized asynchronous algorithms are emerging as a potential alternative but their practical applicability still lags. In order to mitigate the increase in communication cost that naturally comes with scaling the number of workers, we introduce a principled asynchronous, randomized, gossip-based optimization algorithm which works thanks to a continuous local momentum named $\textbf{A}^2\textbf{CiD}^2$. Our method allows each worker to continuously process mini-batches without stopping, and run a peer-to-peer averaging routine in parallel, reducing idle time. In addition to inducing a significant communication acceleration at no cost other than adding a local momentum variable, minimal adaptation is required to incorporate $\textbf{A}^2\textbf{CiD}^2$ to standard asynchronous approaches. Our theoretical analysis proves accelerated rates compared to previous asynchronous decentralized baselines and we empirically show that using our $\textbf{A}^2\textbf{CiD}^2$ momentum significantly decrease communication costs in poorly connected networks. In particular, we show consistent improvement on the ImageNet dataset using up to 64 asynchronous workers (A100 GPUs) and various communication network topologies.

Gaussian processes are used in many machine learning applications that rely on uncertainty quantification. Recently, computational tools for working with these models in geometric settings, such as when inputs lie on a Riemannian manifold, have been developed. This raises the question: can these intrinsic models be shown theoretically to lead to better performance, compared to simply embedding all relevant quantities into $\mathbb{R}^d$ and using the restriction of an ordinary Euclidean Gaussian process? To study this, we prove optimal contraction rates for intrinsic Matérn Gaussian processes defined on compact Riemannian manifolds. We also prove analogous rates for extrinsic processes using trace and extension theorems between manifold and ambient Sobolev spaces: somewhat surprisingly, the rates obtained turn out to coincide with those of the intrinsic processes, provided that their smoothness parameters are matched appropriately. We illustrate these rates empirically on a number of examples, which, mirroring prior work, show that intrinsic processes can achieve better performance in practice. Therefore, our work shows that finer-grained analyses are needed to distinguish between different levels of data-efficiency of geometric Gaussian processes, particularly in settings which involve small data set sizes and non-asymptotic behavior.

Pseudo labeling (PL) is a wide-applied strategy to enlarge the labeled dataset by self-annotating the potential samples during the training process. Several works have shown that it can improve the graph learning model performance in general. However, we notice that the incorrect labels can be fatal to the graph training process. Inappropriate PL may result in the performance degrading, especially on graph data where the noise can propagate. Surprisingly, the corresponding error is seldom theoretically analyzed in the literature. In this paper, we aim to give deep insights of PL on graph learning models. We first present the error analysis of PL strategy by showing that the error is bounded by the confidence of PL threshold and consistency of multi-view prediction. Then, we theoretically illustrate the effect of PL on convergence property. Based on the analysis, we propose a cautious pseudo labeling methodology in which we pseudo label the samples with highest confidence and multi-view consistency. Finally, extensive experiments demonstrate that the proposed strategy improves graph learning process and outperforms other PL strategies on link prediction and node classification tasks.

Robustness to distribution shift has become a growing concern for text and image models as they transition from research subjects to deployment in the real world. However, high-quality benchmarks for distribution shift in tabular machine learning tasks are still lacking despite the widespread real-world use of tabular data and differences in the models used for tabular data in comparison to text and images. As a consequence, the robustness of tabular models to distribution shift is poorly understood. To address this issue, we introduce TableShift, a distribution shift benchmark for tabular data. TableShift contains 15 binary classification tasks in total, each with an associated shift, and includes a diverse set of data sources, prediction targets, and distribution shifts. The benchmark covers domains including finance, education, public policy, healthcare, and civic participation, and is accessible using only a few lines of Python code via the TableShift API. We conduct a large-scale study comparing several state-of-the-art tabular data models alongside robust learning and domain generalization methods on the benchmark tasks. Our study demonstrates (1) a linear trend between in-distribution (ID) and out-of-distribution (OOD) accuracy; (2) domain robustness methods can reduce shift gaps but at the cost of reduced ID accuracy; (3) a strong relationship between shift gap (difference between ID and OOD performance) and shifts in the label distribution. The benchmark data, Python package, model implementations, and more information about TableShift are available at https://github.com/mlfoundations/tableshift and https://tableshift.org .

Training data attribution (TDA) techniques find influential training data for the model's prediction on the test data of interest. They approximate the impact of down- or up-weighting a particular training sample. While conceptually useful, they are hardly applicable to deep models in practice, particularly because of their sensitivity to different model initialisation. In this paper, we introduce a Bayesian perspective on the TDA task, where the learned model is treated as a Bayesian posterior and the TDA estimates as random variables. From this novel viewpoint, we observe that the influence of an individual training sample is often overshadowed by the noise stemming from model initialisation and SGD batch composition. Based on this observation, we argue that TDA can only be reliably used for explaining deep model predictions that are consistently influenced by certain training data, independent of other noise factors. Our experiments demonstrate the rarity of such noise-independent training-test data pairs but confirm their existence. We recommend that future researchers and practitioners trust TDA estimates only in such cases. Further, we find a disagreement between ground truth and estimated TDA distributions and encourage future work to study this gap. Code is provided at https://github.com/ElisaNguyen/bayesian-tda.

Table Representation Learning (TRL) models are commonly pre-trained on large open-domain datasets comprising millions of tables and then used to address downstream tasks. Choosing the right TRL model to use on proprietary data can be challenging, as the best results depend on the content domain, schema, and data quality. Our purpose is to support end-users in testing TRL models on proprietary data in two established SQL-centric tasks, i.e., Question Answering (QA) and Semantic Parsing (SP). We present QATCH (Query-Aided TRL Checklist), a toolbox to highlight TRL models’ strengths and weaknesses on relational tables unseen at training time. For an input table, QATCH automatically generates a testing checklist tailored to QA and SP. Checklist generation is driven by a SQL query engine that crafts tests of different complexity. This design facilitates inherent portability, allowing the checks to be used by alternative models. We also introduce a set of cross-task performance metrics evaluating the TRL model’s performance over its output. Finally, we show how QATCH automatically generates tests for proprietary datasets to evaluate various state-of-the-art models including TAPAS, TAPEX, and CHATGPT.

Deep learning methods have achieved state-of-the-art performance in most modeling tasks involving images, text and audio, however, they typically underperform tree-based methods on tabular data. In this paper, we hypothesize that a significant contributor to this performance gap is the interaction between irregular target functions resulting from the heterogeneous nature of tabular feature spaces, and the well-known tendency of neural networks to learn smooth functions. Utilizing tools from spectral analysis, we show that functions described by tabular datasets often have high irregularity, and that they can be smoothed by transformations such as scaling and ranking in order to improve performance. However, because these transformations tend to lose information or negatively impact the loss landscape during optimization, they need to be rigorously fine-tuned for each feature to achieve performance gains. To address these problems, we propose introducing frequency reduction as an inductive bias. We realize this bias as a neural network layer that promotes learning low-frequency representations of the input features, allowing the network to operate in a space where the target function is more regular. Our proposed method introduces less computational complexity than a fully connected layer, while significantly improving neural network performance, and speeding up its convergence on 14 tabular datasets.

Ensembles of machine learning models have been well established as a powerful method of improving performance over a single model. Traditionally, ensembling algorithms train their base learners independently or sequentially with the goal of optimizing their joint performance. In the case of deep ensembles of neural networks, we are provided with the opportunity to directly optimize the true objective: the joint performance of the ensemble as a whole. Surprisingly, however, directly minimizing the loss of the ensemble appears to rarely be applied in practice. Instead, most previous research trains individual models independently with ensembling performed _post hoc_. In this work, we show that this is for good reason - _joint optimization of ensemble loss results in degenerate behavior_. We approach this problem by decomposing the ensemble objective into the strength of the base learners and the diversity between them. We discover that joint optimization results in a phenomenon in which base learners collude to artificially inflate their apparent diversity. This pseudo-diversity fails to generalize beyond the training data, causing a larger generalization gap. We proceed to comprehensively demonstrate the practical implications of this effect on a range of standard machine learning tasks and architectures by smoothly interpolating between independent training and joint optimization.

While personalized recommendations systems have become increasingly popular, ensuring user data protection remains a top concern in the development of these learning systems. A common approach to enhancing privacy involves training models using anonymous data rather than individual data. In this paper, we explore a natural technique called "look-alike clustering", which involves replacing sensitive features of individuals with the cluster's average values. We provide a precise analysis of how training models using anonymous cluster centers affects their generalization capabilities. We focus on an asymptotic regime where the size of the training set grows in proportion to the features dimension. Our analysis is based on the Convex Gaussian Minimax Theorem (CGMT) and allows us to theoretically understand the role of different model components on the generalization error. In addition, we demonstrate that in certain high-dimensional regimes, training over anonymous cluster centers acts as a regularization and improves generalization error of the trained models. Finally, we corroborate our asymptotic theory with finite-sample numerical experiments where we observe a perfect match when the sample size is only of order of a few hundreds.

The use of non-Cartesian grids is a niche but important topic in sub-fields of the numerical sciences such as simulation and scientific visualization. However, non-Cartesian approaches are virtually unexplored in machine learning. This is likely due to the difficulties in the representation of data on non-Cartesian domains and the lack of support for standard machine learning operations on non-Cartesian data. This paper proposes a new data structure called the lattice tensor which generalizes traditional tensor spatio-temporal operations to lattice tensors, enabling the use of standard machine learning algorithms on non-Cartesian data. However, data need not reside on a non-Cartesian structure, we use non-Dyadic downsampling schemes to bring Cartesian data into a non-Cartesian space for further processing. We introduce a software library that implements the lattice tensor container (with some common machine learning operations), and demonstrate its effectiveness. Our method provides a general framework for machine learning on non-Cartesian domains, addressing the challenges mentioned above and filling a gap in the current literature.

Visual object tracking (VOT) is one of the most fundamental tasks in computer vision community. State-of-the-art VOT trackers extract positive and negative examples that are used to guide the tracker to distinguish the object from the background. In this paper, we show that this characteristic can be exploited to introduce new threats and hence propose a simple yet effective poison-only backdoor attack. To be specific, we poison a small part of the training data by attaching a predefined trigger pattern to the background region of each video frame, so that the trigger appears almost exclusively in the extracted negative examples. To the best of our knowledge, this is the first work that reveals the threat of poison-only backdoor attack on VOT trackers. We experimentally show that our backdoor attack can significantly degrade the performance of both two-stream Siamese and one-stream Transformer trackers on the poisoned data while gaining comparable performance with the benign trackers on the clean data.

Generative data augmentation, which scales datasets by obtaining fake labeled examples from a trained conditional generative model, boosts classification performance in various learning tasks including (semi-)supervised learning, few-shot learning, and adversarially robust learning. However, little work has theoretically investigated the effect of generative data augmentation. To fill this gap, we establish a general stability bound in this not independently and identically distributed (non-i.i.d.) setting, where the learned distribution is dependent on the original train set and generally not the same as the true distribution. Our theoretical result includes the divergence between the learned distribution and the true distribution. It shows that generative data augmentation can enjoy a faster learning rate when the order of divergence term is $o(\max\left( \log(m)\beta_m, 1 / \sqrt{m})\right)$, where $m$ is the train set size and $\beta_m$ is the corresponding stability constant. We further specify the learning setup to the Gaussian mixture model and generative adversarial nets. We prove that in both cases, though generative data augmentation does not enjoy a faster learning rate, it can improve the learning guarantees at a constant level when the train set is small, which is significant when the awful overfitting occurs. Simulation results on the Gaussian mixture model and empirical results on generative adversarial nets support our theoretical conclusions.

Machine learning models are famously vulnerable to adversarial attacks: small ad-hoc perturbations of the data that can catastrophically alter the model predictions. While a large literature has studied the case of test-time attacks on pre-trained models, the important case of attacks in an online learning setting has received little attention so far. In this work, we use a control-theoretical perspective to study the scenario where an attacker may perturb data labels to manipulate the learning dynamics of an online learner. We perform a theoretical analysis of the problem in a teacher-student setup, considering different attack strategies, and obtaining analytical results for the steady state of simple linear learners. These results enable us to prove that a discontinuous transition in the learner's accuracy occurs when the attack strength exceeds a critical threshold. We then study empirically attacks on learners with complex architectures using real data, confirming the insights of our theoretical analysis. Our findings show that greedy attacks can be extremely efficient, especially when data stream in small batches.