ICLR 2023 Papers

When trained on language data, do transformers learn some arbitrary computation that utilizes the full capacity of the architecture or do they learn a simpler, tree-like computation, hypothesized to underlie compositional meaning systems like human languages? There is an apparent tension between compositional accounts of human language understanding, which are based on a restricted bottom-up computational process, and the enormous success of neural models like transformers, which can route information arbitrarily between different parts of their input. One possibility is that these models, while extremely flexible in principle, in practice learn to interpret language hierarchically, ultimately building sentence representations close to those predictable by a bottom-up, tree-structured model. To evaluate this possibility, we describe an unsupervised and parameter-free method to \emph{functionally project} the behavior of any transformer into the space of tree-structured networks. Given an input sentence, we produce a binary tree that approximates the transformer's representation-building process and a score that captures how ``tree-like'' the transformer's behavior is on the input. While calculation of this score does not require training any additional models, it provably upper-bounds the fit between a transformer and any tree-structured approximation. Using this method, we show that transformers for three different tasks become more tree-like over the course of training, in some cases unsupervisedly recovering the same trees as supervised parsers. These trees, in turn, are predictive of model behavior, with more tree-like models generalizing better on tests of compositional generalization.

Deep neural networks (DNNs) are nowadays ubiquitous in many domains such as computer vision. However, due to their high latency, the deployment of DNNs hinges on the development of compression techniques such as quantization which consists in lowering the number of bits used to encode the weights and activations. Growing concerns for privacy and security have motivated the development of data-free techniques, at the expanse of accuracy. In this paper, we identity the uniformity of the quantization operator as a limitation of existing approaches, and propose a data-free non-uniform method. More specifically, we argue that to be readily usable without dedicated hardware and implementation, non-uniform quantization shall not change the nature of the mathematical operations performed by the DNN. This leads to search among the continuous automorphisms of $(\mathbb{R}_+^*,\times)$, which boils down to the power functions defined by their exponent. To find this parameter, we propose to optimize the reconstruction error of each layer: in particular, we show that this procedure is locally convex and admits a unique solution. At inference time, we show that our approach, dubbed PowerQuant, only require simple modifications in the quantized DNN activation functions. As such, with only negligible overhead, it significantly outperforms existing methods in a variety of configurations.

Neural Architecture Search (NAS) is widely used to automatically obtain the neural network with the best performance among a large number of candidate architectures. To reduce the search time, zero-shot NAS aims at designing training-free proxies that can predict the test performance of a given architecture. However, as shown recently, none of the zero-shot proxies proposed to date can actually work consistently better than a naive proxy, namely, the number of network parameters (#Params). To improve this state of affairs, as the main theoretical contribution, we first reveal how some specific gradient properties across different samples impact the convergence rate and generalization capacity of neural networks. Based on this theoretical analysis, we propose a new zero-shot proxy, ZiCo, the first proxy that works consistently better than #Params. We demonstrate that ZiCo works better than State-Of-The-Art (SOTA) proxies on several popular NAS-Benchmarks (NASBench101, NATSBench-SSS/TSS, TransNASBench-101) for multiple applications (e.g., image classification/reconstruction and pixel-level prediction). Finally, we demonstrate that the optimal architectures found via ZiCo are as competitive as the ones found by one-shot and multi-shot NAS methods, but with much less search time. For example, ZiCo-based NAS can find optimal architectures with 78.1%, 79.4%, and 80.4% test accuracy under inference budgets of 450M, 600M, and 1000M FLOPs, respectively, on ImageNet within 0.4 GPU days. Our code is available at https://github.com/SLDGroup/ZiCo.

Mixed-integer linear programming (MILP) is widely employed for modeling combinatorial optimization problems. In practice, similar MILP instances with only coefficient variations are routinely solved, and machine learning (ML) algorithms are capable of capturing common patterns across these MILP instances. In this work, we combine ML with optimization and propose a novel predict-and-search framework for efficiently identifying high-quality feasible solutions. Specifically, we first utilize graph neural networks to predict the marginal probability of each variable, and then search for the best feasible solution within a properly defined ball around the predicted solution. We conduct extensive experiments on public datasets, and computational results demonstrate that our proposed framework achieves 51.1% and 9.9% performance improvements to MILP solvers SCIP and Gurobi on primal gaps, respectively.

Current abstractive summarization models either suffer from a lack of clear interpretability or provide incomplete rationales by only highlighting parts of the source document. To this end, we propose the Summarization Program (SP), an interpretable modular framework consisting of an (ordered) list of binary trees, each encoding the step-by-step generative process of an abstractive summary sentence from the source document. A Summarization Program contains one root node per summary sentence, and a distinct tree connects each summary sentence (root node) to the document sentences (leaf nodes) from which it is derived, with the connecting nodes containing intermediate generated sentences. Edges represent different modular operations involved in summarization such as sentence fusion, compression, and paraphrasing. We first propose an efficient best-first search method over neural modules, SP-Search that identifies SPs for human summaries by directly optimizing for ROUGE scores. Next, using these programs as automatic supervision, we propose seq2seq models that generate Summarization Programs, which are then executed to obtain final summaries. We demonstrate that SP-Search effectively represents the generative process behind human summaries using modules that are typically faithful to their intended behavior. We also conduct a simulation study to show that Summarization Programs improve the interpretability of summarization models by allowing humans to better simulate model reasoning. Summarization Programs constitute a promising step toward interpretable and modular abstractive summarization, a complex task previously addressed primarily through blackbox end-to-end neural systems.

The energy consumption for training deep learning models is increasing at an alarming rate due to the growth of training data and model scale, resulting in a negative impact on carbon neutrality. Energy consumption is an especially pressing issue for AutoML algorithms because it usually requires repeatedly training large numbers of computationally intensive deep models to search for optimal configurations. This paper takes one of the most essential steps in developing energy-aware (EA) NAS methods, by providing a benchmark that makes EA-NAS research more reproducible and accessible. Specifically, we present the first large-scale energy-aware benchmark that allows studying AutoML methods to achieve better trade-offs between performance and search energy consumption, named EA-HAS-Bench. EA-HAS-Bench provides a large-scale architecture/hyperparameter joint search space, covering diversified configurations related to energy consumption. Furthermore, we propose a novel surrogate model specially designed for large joint search space, which proposes a Bezier curve-based model to predict learning curves with unlimited shape and length. Based on the proposed dataset, we new energy-aware AutoML method that arms existing AutoML algorithms to consider the search energy consumption, and our experiments show that the modified energy-aware AutoML methods achieve a better trade-off between energy consumption and model performance.

This paper asks the intriguing question: is it possible to exploit neural architecture search (NAS) as a new attack vector to launch previously improbable attacks? Specifically, we present EVAS, a new attack that leverages NAS to find neural architectures with inherent backdoors and exploits such vulnerability using input-aware triggers. Compared with existing attacks, EVAS demonstrates many interesting properties: (i) it does not require polluting training data or perturbing model parameters; (ii) it is agnostic to downstream fine-tuning or even re-training from scratch; (iii) it naturally evades defenses that rely on inspecting model parameters or training data. With extensive evaluation on benchmark datasets, we show that EVAS features high evasiveness, transferability, and robustness, thereby expanding the adversary's design spectrum. We further characterize the mechanisms underlying EVAS, which are possibly explainable by architecture-level ``shortcuts'' that recognize trigger patterns. This work showcases that NAS can be exploited in a harmful way to find architectures with inherent backdoor vulnerability. The code is available at https://github.com/ain-soph/nas_backdoor.

A decision tree is a fundamental model used in data mining and machine learning. In practice, the training data used to construct a decision tree may change over time or contain noise, and a drastic change in the learned tree structure owing to such data perturbation is unfavorable. For example, in data mining, a change in the tree implies a change in the extracted knowledge, which raises the question of whether the extracted knowledge is truly reliable or is only a noisy artifact. To alleviate this issue, we design decision tree learning algorithms that are stable against insignificant perturbations in the training data. Specifically, we adopt the notion of average sensitivity as a stability measure, and design an algorithm with low average sensitivity that outputs a decision tree whose accuracy is close to the optimal decision tree. The experimental results on real-world datasets demonstrate that the proposed algorithm enables users to select suitable decision trees considering the trade-off between average sensitivity and accuracy.

Nonlinear dynamics is ubiquitous in nature and commonly seen in various science and engineering disciplines. Distilling analytical expressions that govern nonlinear dynamics from limited data remains vital but challenging. To tackle this fundamental issue, we propose a novel Symbolic Physics Learner (SPL) machine to discover the mathematical structure of nonlinear dynamics. The key concept is to interpret mathematical operations and system state variables by computational rules and symbols, establish symbolic reasoning of mathematical formulas via expression trees, and employ a Monte Carlo tree search (MCTS) agent to explore optimal expression trees based on measurement data. The MCTS agent obtains an optimistic selection policy through the traversal of expression trees, featuring the one that maps to the arithmetic expression of underlying physics. Salient features of the proposed framework include search flexibility and enforcement of parsimony for discovered equations. The efficacy and superiority of the SPL machine are demonstrated by numerical examples, compared with state-of-the-art baselines.

MuZero Unplugged presents a promising approach for offline policy learning from logged data. It conducts Monte-Carlo Tree Search (MCTS) with a learned model and leverages Reanalyze algorithm to learn purely from offline data. For good performance, MCTS requires accurate learned models and a large number of simulations, thus costing huge computing time. This paper investigates a few hypotheses where MuZero Unplugged may not work well under the offline RL settings, including 1) learning with limited data coverage; 2) learning from offline data of stochastic environments; 3) improperly parameterized models given the offline data; 4) with a low compute budget. We propose to use a regularized one-step look-ahead approach to tackle the above issues. Instead of planning with the expensive MCTS, we use the learned model to construct an advantage estimation based on a one-step rollout. Policy improvements are towards the direction that maximizes the estimated advantage with regularization of the dataset. We conduct extensive empirical studies with BSuite environments to verify the hypotheses and then run our algorithm on the RL Unplugged Atari benchmark. Experimental results show that our proposed approach achieves stable performance even with an inaccurate learned model. On the large-scale Atari benchmark, the proposed method outperforms MuZero Unplugged by 43%. Most significantly, it uses only 5.6% wall-clock time (i.e., 1 hour) compared to MuZero Unplugged (i.e., 17.8 hours) to achieve a 150% IQM normalized score with the same hardware and software stacks. Our implementation is open-sourced at https://github.com/sail-sg/rosmo.

A soft tree is an actively studied variant of a decision tree that updates splitting rules using the gradient method. Although soft trees can take various architectures, their impact is not theoretically well known. In this paper, we formulate and analyze the Neural Tangent Kernel (NTK) induced by soft tree ensembles for arbitrary tree architectures. This kernel leads to the remarkable finding that only the number of leaves at each depth is relevant for the tree architecture in ensemble learning with an infinite number of trees. In other words, if the number of leaves at each depth is fixed, the training behavior in function space and the generalization performance are exactly the same across different tree architectures, even if they are not isomorphic. We also show that the NTK of asymmetric trees like decision lists does not degenerate when they get infinitely deep. This is in contrast to the perfect binary trees, whose NTK is known to degenerate and leads to worse generalization performance for deeper trees.

Differentiable neural architecture search methods are increasingly popular due to their computational efficiency. However, these methods have unsatisfactory generalizability and stability. Their searched architectures are often degenerate with a dominant number of skip connections and perform unsatisfactorily on test data. Existing methods for solving this problem have a variety of limitations, such as cannot prevent the happening of architecture degeneration, being excessively restrictive in setting the number of skip connections, etc. To address these limitations, we propose a new approach for improving the generalizability and stability of differentiable NAS, by developing a transferability-encouraging tri-level optimization framework which improves the architecture of a main model by encouraging good transferability to an auxiliary model. Our framework involves three stages performed end-to-end: 1) train network weights of a main model; 2) transfer knowledge from the main model to an auxiliary model; 3) optimize the architecture of the main model by maximizing its transferability to the auxiliary model. We propose a new knowledge transfer approach based on matching quadruple relative similarities. Experiments on several datasets demonstrate the effectiveness of our method.

We introduce CriticSMC, a new algorithm for planning as inference built from a composition of sequential Monte Carlo with learned Soft-Q function heuristic factors. These heuristic factors, obtained from parametric approximations of the marginal likelihood ahead, more effectively guide SMC towards the desired target distribution, which is particularly helpful for planning in environments with hard constraints placed sparsely in time. Compared with previous work, we modify the placement of such heuristic factors, which allows us to cheaply propose and evaluate large numbers of putative action particles, greatly increasing inference and planning efficiency. CriticSMC is compatible with informative priors, whose density function need not be known, and can be used as a model-free control algorithm. Our experiments on collision avoidance in a high-dimensional simulated driving task show that CriticSMC significantly reduces collision rates at a low computational cost while maintaining realism and diversity of driving behaviors across vehicles and environment scenarios.

Face detection (FD) has achieved remarkable success over the past few years, yet, these leaps often arrive when consuming enormous computation costs. Moreover, when considering a realistic situation, i.e., building a lightweight face detector under a computation-scarce scenario, such heavy computation cost limits the application of the face detector. To remedy this, several pioneering works design tiny face detectors through off-the-shelf neural architecture search (NAS) technologies, which are usually applied to the classification task. Thus, the searched architectures are sub-optimal for the face detection task since some design criteria between detection and classification task are different. As a representative, the face detection backbone design needs to guarantee the stage-level detection ability while it is not required for the classification backbone. Furthermore, the detection backbone consumes a vast body of inference budgets in the whole detection framework. Considering the intrinsic design requirement and the virtual importance role of the face detection backbone, we thus ask a critical question: How to employ NAS to search FD-friendly backbone architecture? To cope with this question, we propose a distribution-dependent stage-aware ranking score (DDSAR-Score) to explicitly characterize the stage-level expressivity and identify the individual importance of each stage, thus satisfying the aforementioned design criterion of the FD backbone. Based on our proposed DDSAR-Score, we conduct comprehensive experiments on the challenging Wider Face benchmark dataset and achieve dominant performance across a wide range of compute regimes. In particular, compared to the tiniest face detector SCRFD-0.5GF, our method is +2.5 % better in Average Precision (AP) score when using the same amount of FLOPs. The code is avaliable at https://github.com/ly19965/FaceMaas/tree/master/face_project/face_detection/DamoFD.

The approximate nearest neighbor (ANN) search problem is fundamental to efficiently serving many real-world machine learning applications. A number of techniques have been developed for ANN search that are efficient, accurate, and scalable. However, such techniques typically have a number of parameters that affect the speed-recall tradeoff, and exhibit poor performance when such parameters aren't properly set. Tuning these parameters has traditionally been a manual process, demanding in-depth knowledge of the underlying search algorithm. This is becoming an increasingly unrealistic demand as ANN search grows in popularity. To tackle this obstacle to ANN adoption, this work proposes a constrained optimization-based approach to tuning quantization-based ANN algorithms. Our technique takes just a desired search cost or recall as input, and then generates tunings that, empirically, are very close to the speed-recall Pareto frontier and give leading performance on standard benchmarks.

Given an unlabeled pool of data and the experts who can label them, active learning aims to build an agent that can effectively acquire data to be queried to the experts, maximizing the gain in performance when trained with them. While there are several principles for active learning, a prevailing approach is to estimate uncertainties of predictions for unlabeled samples and use them to define acquisition functions. Active learning with the uncertainty principle works well for deep learning, especially for large-scale image classification tasks with deep neural networks. Still, it is often overlooked how the uncertainty of predictions is estimated, despite the common findings on the difficulty of accurately estimating uncertainties of deep neural networks. In this paper, we highlight the effectiveness of snapshot ensembles for deep active learning. Compared to the previous approaches based on Monte-Carlo dropout or deep ensembles, we show that a simple acquisition strategy based on uncertainties estimated from parameter snapshots gathered from a single optimization path significantly improves the quality of the acquired samples. Based on this observation, we further propose an efficient active learning algorithm that maintains a single learning trajectory throughout the entire active learning episodes, unlike the existing algorithms training models from scratch for every active learning episode. Through the extensive empirical comparison, we demonstrate the effectiveness of snapshot ensembles for deep active learning.

Although Transformer architectures have been successfully applied to graph data with the advent of Graph Transformer, current design of Graph Transformer still heavily relies on human labor and expertise knowledge to decide proper neural architectures and suitable graph encoding strategies at each Transformer layer. In literature, there have been some works on automated design of Transformers focusing on non-graph data such as texts and images without considering graph encoding strategies, which fail to handle the non-euclidean graph data. In this paper, we study the problem of automated graph Transformer, for the first time. However, solving these problems poses the following challenges: i) how can we design a unified search space for graph Transformer, and ii) how to deal with the coupling relations between Transformer architectures and the graph encodings of each Transformer layer. To address these challenges, we propose Automated Graph Transformer (AutoGT), a neural architecture search framework that can automatically discover the optimal graph Transformer architectures by joint optimization of Transformer architecture and graph encoding strategies. Specifically, we first propose a unified graph Transformer formulation that can represent most of state-of-the-art graph Transformer architectures. Based upon the unified formulation, we further design the graph Transformer search space that includes both candidate architectures and various graph encodings. To handle the coupling relations, we propose a novel encoding-aware performance estimation strategy by gradually training and splitting the supernets according to the correlations between graph encodings and architectures. The proposed strategy can provide a more consistent and fine-grained performance prediction when evaluating the jointly optimized graph encodings and architectures. Extensive experiments and ablation studies show that our proposed AutoGT gains sufficient improvement over state-of-the-art hand-crafted baselines on all datasets, demonstrating its effectiveness and wide applicability.

Hamiltonian Monte Carlo (HMC) is a popular method in sampling. While there are quite a few works of studying this method on various aspects, an interesting question is how to choose its integration time to achieve acceleration. In this work, we consider accelerating the process of sampling from a distribution $\pi(x) \propto \exp(-f(x))$ via HMC via time-varying integration time. When the potential $f$ is $L$-smooth and $m$-strongly convex, i.e. for sampling from a log-smooth and strongly log-concave target distribution $\pi$, it is known that under a constant integration time, the number of iterations that ideal HMC takes to get an $\epsilon$ Wasserstein-2 distance to the target $\pi$ is $O( \kappa \log \frac{1}{\epsilon} )$, where $\kappa := \frac{L}{m}$ is the condition number. We propose a scheme of time-varying integration time based on the roots of Chebyshev polynomials. We show that in the case of quadratic potential $f$, i.e. when the target $\pi$ is a Gaussian distribution, ideal HMC with this choice of integration time only takes $O( \sqrt{\kappa} \log \frac{1}{\epsilon} )$ number of iterations to reach Wasserstein-2 distance less than $\epsilon$; this improvement on the dependence on condition number is akin to acceleration in optimization. The design and analysis of HMC with the proposed integration time is built on the tools of Chebyshev polynomials. Experiments find the advantage of adopting our scheme of time-varying integration time even for sampling from distributions with smooth strongly convex potentials that are not quadratic.

Denoising diffusion models are a popular class of generative models providing state-of-the-art results in many domains. One adds gradually noise to data using a diffusion to transform the data distribution into a Gaussian distribution. Samples from the generative model are then obtained by simulating an approximation of the time-reversal of this diffusion initialized by Gaussian samples. Practically, the intractable score terms appearing in the time-reversed process are approximated using score matching techniques. We explore here a similar idea to sample approximately from unnormalized probability density functions and estimate their normalizing constants. We consider a process where the target density diffuses towards a Gaussian. Denoising Diffusion Samplers (DDS) are obtained by approximating the corresponding time-reversal. While score matching is not applicable in this context, we can leverage many of the ideas introduced in generative modeling for Monte Carlo sampling. Existing theoretical results from denoising diffusion models also provide theoretical guarantees for DDS. We discuss the connections between DDS, optimal control and Schr\"odinger bridges and finally demonstrate DDS experimentally on a variety of challenging sampling tasks.