_id stringlengths 4 10 | text stringlengths 0 18.4k | title stringlengths 0 8.56k |
|---|---|---|
d219981518 | Calibrating neural networks is of utmost importance when employing them in safety-critical applications where the downstream decision making depends on the predicted probabilities. Measuring calibration error amounts to comparing two empirical distributions. In this work, we introduce a binning-free calibration measure... | Published as a conference paper at ICLR 2021 CALIBRATION OF NEURAL NETWORKS USING SPLINES |
d3543617 | Deep generative models have been enjoying success in modeling continuous data. However it remains challenging to capture the representations for discrete structures with formal grammars and semantics, e.g., computer programs and molecular structures. How to generate both syntactically and semantically correct data stil... | Published as a conference paper at ICLR 2018 SYNTAX-DIRECTED VARIATIONAL AUTOENCODER FOR STRUCTURED DATA |
d256615568 | Moiré patterns appear frequently when taking photos of digital screens, drastically degrading the image quality. Despite the advance of CNNs in image demoiréing, existing networks are with heavy design, causing redundant computation burden for mobile devices. In this paper, we launch the first study on accelerating dem... | REAL-TIME IMAGE DEMOIRÉING ON MOBILE DE- VICES |
d249375525 | The asymptotic mean squared test error and sensitivity of the Random Features Regression model (RFR) have been recently studied. We build on this work and identify in closed-form the family of Activation Functions (AFs) that minimize a combination of the test error and sensitivity of the RFR under different notions of ... | Published as a conference paper at ICLR 2023 OPTIMAL ACTIVATION FUNCTIONS FOR THE RANDOM FEATURES REGRESSION MODEL |
d15530352 | Dataset augmentation, the practice of applying a wide array of domain-specific transformations to synthetically expand a training set, is a standard tool in supervised learning. While effective in tasks such as visual recognition, the set of transformations must be carefully designed, implemented, and tested for every ... | Workshop track -ICLR 2017 DATASET AUGMENTATION IN FEATURE SPACE |
d244478279 | The lottery ticket hypothesis has sparked the rapid development of pruning algorithms that aim to reduce the computational costs associated with deep learning during training and model deployment. Currently, such algorithms are primarily evaluated on imaging data, for which we lack ground truth information and thus the... | Published as a conference paper at ICLR 2022 PLANT 'N' SEEK: CAN YOU FIND THE WINNING TICKET? |
d258437184 | Machine learning models fail to perform when facing out-of-distribution (OOD) domains, a challenging task known as domain generalization (DG). In this work, we develop a novel DG training strategy, we call PGrad , to learn a robust gradient direction, improving models' generalization ability on unseen domains. The prop... | Published as a conference paper at ICLR 2023 PGR A D : LEARNING PRINCIPAL GRADIENTS FOR DO- MAIN GENERALIZATION |
d252199584 | Partial differential equations (PDEs) see widespread use in sciences and engineering to describe simulation of physical processes as scalar and vector fields interacting and coevolving over time. Due to the computationally expensive nature of their standard solution methods, neural PDE surrogates have become an active ... | Published as a conference paper at ICLR 2023 CLIFFORD NEURAL LAYERS FOR PDE MODELING |
d235266229 | In recent years, implicit deep learning has emerged as a method to increase the effective depth of deep neural networks. While their training is memory-efficient, they are still significantly slower to train than their explicit counterparts. In Deep Equilibrium Models (DEQs), the training is performed as a bi-level pro... | SHINE: SHARING THE INVERSE ESTIMATE FROM THE FORWARD PASS FOR BI-LEVEL OPTIMIZATION AND IM- PLICIT MODELS |
d236034533 | Recent progress in language model pre-training has achieved a great success via leveraging large-scale unstructured textual data. However, it is still a challenge to apply pre-training on structured tabular data due to the absence of large-scale high-quality tabular data. In this paper, we propose TAPEX to show that ta... | Published as a conference paper at ICLR 2022 TAPEX: TABLE PRE-TRAINING VIA LEARNING A NEURAL SQL EXECUTOR |
d252118781 | We present a neural network architecture, Bispectral Neural Networks (BNNs) for learning representations that are invariant to the actions of compact commutative groups on the space over which a signal is defined. The model incorporates the ansatz of the bispectrum, an analytically defined group invariant that is compl... | Published as a conference paper at ICLR 2023 BISPECTRAL NEURAL NETWORKS |
d234742529 | We consider the problem of learning a latent -vertex simplex ⊂ ℝ , given access to A ∈ ℝ × , which can be viewed as a data matrix with points that are obtained by randomly perturbing latent points in the simplex (potentially beyond ). A large class of latent variable models, such as adversarial clustering, mixed member... | Learning a Latent Simplex in Input-Sparsity Time |
d218487350 | Although Neural Differential Equations have shown promise on toy problems such as MNIST, they have yet to be successfully applied to more challenging tasks. Inspired by variational methods for image restoration relying on partial differential equations, we choose to benchmark several forms of Neural DEs and backpropaga... | Neural Differential Equations for Single Image Super-Resolution |
d211259530 | The vulnerabilities of deep neural networks against adversarial examples have become a significant concern for deploying these models in sensitive domains. Devising a definitive defense against such attacks is proven to be challenging, and the methods relying on detecting adversarial samples are only valid when the att... | ADVERSARIAL EXAMPLE DETECTION AND CLASSIFI- CATION WITH ASYMMETRICAL ADVERSARIAL TRAIN- ING |
d259108315 | Though Self-supervised learning (SSL) has been widely studied as a promising technique for representation learning, it doesn't generalize well on long-tailed datasets due to the majority classes dominating the feature space. Recent work shows that the long-tailed learning performance could be boosted by sampling extra ... | ON THE EFFECTIVENESS OF OUT-OF-DISTRIBUTION DATA IN SELF-SUPERVISED LONG-TAIL LEARNING |
d485828 | We propose an approach to learn spatio-temporal features in videos from intermediate visual representations we call "percepts" using Gated-Recurrent-Unit Recurrent Networks (GRUs). Our method relies on percepts that are extracted from all levels of a deep convolutional network trained on the large ImageNet dataset. Whi... | Published as a conference paper at ICLR 2016 DELVING DEEPER INTO CONVOLUTIONAL NETWORKS FOR LEARNING VIDEO REPRESENTATIONS |
d257102785 | We tackle the domain generalisation (DG) problem by posing it as a domain adaptation (DA) task where we adversarially synthesise the worst-case 'target' domain and adapt a model to that worst-case domain, thereby improving the model's robustness. To synthesise data that is challenging yet semantics-preserving, we gener... | Published as a conference paper at ICLR 2023 DOMAIN GENERALISATION VIA DOMAIN ADAPTA- TION: AN ADVERSARIAL FOURIER AMPLITUDE AP- PROACH |
d1597636 | Biologically inspired, from the early HMAX model to Spatial Pyramid Matching, pooling has played an important role in visual recognition pipelines. Spatial pooling, by grouping of local codes, equips these methods with a certain degree of robustness to translation and deformation yet preserving important spatial inform... | Learnable Pooling Regions for Image Classification |
d220128149 | Adversarial poisoning attacks distort training data in order to corrupt the test-time behavior of a classifier. A provable defense provides a certificate for each test sample, which is a lower bound on the magnitude of any adversarial distortion of the training set that can corrupt the test sample's classification. We ... | Deep Partition Aggregation: Provable Defense against General Poisoning Attacks |
d219981806 | A key challenge in adversarial robustness is the lack of a precise mathematical characterization of human perception, used in the definition of adversarial attacks that are imperceptible to human eyes. Most current attacks and defenses try to avoid this issue by considering restrictive adversarial threat models such as... | Published as a conference paper at ICLR 2021 PERCEPTUAL ADVERSARIAL ROBUSTNESS: DEFENSE AGAINST UNSEEN THREAT MODELS |
d59316418 | Transfer learning through fine-tuning a pre-trained neural network with an extremely large dataset, such as ImageNet, can significantly accelerate training while the accuracy is frequently bottlenecked by the limited dataset size of the new target task. To solve the problem, some regularization methods, constraining th... | DELTA: DEEP LEARNING TRANSFER USING FEATURE MAP WITH ATTENTION FOR CONVOLUTIONAL NET- WORKS |
d232240622 | Model-agnostic meta-learning (MAML) is a popular method for few-shot learning but assumes that we have access to the meta-training set. In practice, training on the meta-training set may not always be an option due to data privacy concerns, intellectual property issues, or merely lack of computing resources. In this pa... | Published as a conference paper at ICLR 2021 REPURPOSING PRETRAINED MODELS FOR ROBUST OUT-OF-DOMAIN FEW-SHOT LEARNING |
d237291550 | With recent progress in joint modeling of visual and textual representations, Vision-Language Pretraining (VLP) has achieved impressive performance on many multimodal downstream tasks. However, the requirement for expensive annotations including clean image captions and regional labels limits the scalability of existin... | SIMVLM: SIMPLE VISUAL LANGUAGE MODEL PRE- TRAINING WITH WEAK SUPERVISION |
d7953396 | In recent years, a lot of attention has been devoted to efficient nearest neighbor search by means of similarity-preserving hashing. One of the plights of existing hashing techniques is the intrinsic trade-off between performance and computational complexity: while longer hash codes allow for lower false positive rates... | Sparse similarity-preserving hashing |
d257050560 | Skip connections and normalisation layers form two standard architectural components that are ubiquitous for the training of Deep Neural Networks (DNNs), but whose precise roles are poorly understood. Recent approaches such as Deep Kernel Shaping have made progress towards reducing our reliance on them, using insights ... | Published as a conference paper at ICLR 2023 DEEP TRANSFORMERS WITHOUT SHORTCUTS: MODIFYING SELF-ATTENTION FOR FAITHFUL SIGNAL PROPAGATION |
d251067024 | Proximal splitting algorithms are well suited to solving large-scale nonsmooth optimization problems, in particular those arising in machine learning. We propose a new primal-dual algorithm, in which the dual update is randomized; equivalently, the proximity operator of one of the function in the problem is replaced by... | RANDPROX: PRIMAL-DUAL OPTIMIZATION ALGO- RITHMS WITH RANDOMIZED PROXIMAL UPDATES |
d254853899 | Deep latent variable models have achieved significant empirical successes in modelbased reinforcement learning (RL) due to their expressiveness in modeling complex transition dynamics. On the other hand, it remains unclear theoretically and empirically how latent variable models may facilitate learning, planning, and e... | Published as a conference paper at ICLR 2023 LATENT VARIABLE REPRESENTATION FOR REIN- FORCEMENT LEARNING |
d2835189 | Neuromorphic hardware tends to pose limits on the connectivity of deep networks that one can run on them. But also generic hardware and software implementations of deep learning run more efficiently for sparse networks. Several methods exist for pruning connections of a neural network after it was trained without conne... | DEEP REWIRING: TRAINING VERY SPARSE DEEP NET- WORKS |
d256194594 | Self-supervised pretraining has been extensively studied in language and vision domains, where a unified model can be easily adapted to various downstream tasks by pretraining representations without explicit labels. When it comes to sequential decision-making tasks, however, it is difficult to properly design such a p... | Published as a conference paper at ICLR 2023 SMART: SELF-SUPERVISED MULTI-TASK PRETRAIN- ING WITH CONTROL TRANSFORMERS |
d222291295 | To alleviate the resource constraint for real-time point cloud applications that run on edge devices, in this paper we present BiPointNet, the first model binarization approach for efficient deep learning on point clouds. We discover that the immense performance drop of binarized models for point clouds mainly stems fr... | Published as a conference paper at ICLR 2021 BIPOINTNET: BINARY NEURAL NETWORK FOR POINT CLOUDS |
d257279896 | We present a deep learning approach for repairing sequential circuits against formal specifications given in linear-time temporal logic (LTL). Given a defective circuit and its formal specification, we train Transformer models to output circuits that satisfy the corresponding specification. We propose a separated hiera... | ITERATIVE CIRCUIT REPAIR AGAINST FORMAL SPEC- IFICATIONS |
d221819379 | Multi-Task Learning (MTL) networks have emerged as a promising method for transferring learned knowledge across different tasks. However, MTL must deal with challenges such as: overfitting to low resource tasks, catastrophic forgetting, and negative task transfer, or learning interference. Often, in Natural Language Pr... | Published as a conference paper at ICLR 2021 CONDITIONALLY ADAPTIVE MULTI-TASK LEARNING: IMPROVING TRANSFER LEARNING IN NLP USING FEWER PARAMETERS & LESS DATA |
d221376381 | We construct an experimental setup in which changing the scale of initialization strongly impacts the implicit regularization induced by SGD, interpolating from good generalization performance to completely memorizing the training set while making little progress on the test set. Moreover, we find that the extent and m... | EXTREME MEMORIZATION VIA SCALE OF INITIALIZA- TION |
d3875075 | We present a formal language with expressions denoting general symbol structures and queries which access information in those structures. A sequence-to-sequence network processing this language learns to encode symbol structures and query them. The learned representation (approximately) shares a simple linearity prope... | Workshop track -ICLR 2018 LEARNING AND ANALYZING VECTOR ENCODING OF SYMBOLIC REPRESENTATIONS |
d170078603 | While tasks could come with varying the number of instances and classes in realistic settings, the existing meta-learning approaches for few-shot classification assume that the number of instances per task and class is fixed. Due to such restriction, they learn to equally utilize the meta-knowledge across all the tasks... | Published as a conference paper at ICLR 2020 LEARNING TO BALANCE: BAYESIAN META-LEARNING FOR IMBALANCED AND OUT-OF-DISTRIBUTION TASKS |
d11553675 | Automatic speech recognition systems usually rely on spectral-based features, such as MFCC of PLP. These features are extracted based on prior knowledge such as, speech perception or/and speech production. Recently, convolutional neural networks have been shown to be able to estimate phoneme conditional probabilities i... | Under review as a conference paper at ICLR 2015 LEARNING LINEARLY SEPARABLE FEATURES FOR SPEECH RECOGNITION USING CONVOLUTIONAL NEU- RAL NETWORKS |
d6212000 | Despite their massive size, successful deep artificial neural networks can exhibit a remarkably small difference between training and test performance. Conventional wisdom attributes small generalization error either to properties of the model family, or to the regularization techniques used during training. Through ex... | UNDERSTANDING DEEP LEARNING REQUIRES RE- THINKING GENERALIZATION |
d25717172 | The driving force behind the recent success of LSTMs has been their ability to learn complex and non-linear relationships. Consequently, our inability to describe these relationships has led to LSTMs being characterized as black boxes. To this end, we introduce contextual decomposition (CD), an interpretation algorithm... | Under review as a conference paper at ICLR 2018 BEYOND WORD IMPORTANCE: CONTEXTUAL DE- COMPOSITION TO EXTRACT INTERACTIONS FROM LSTMS |
d209862859 | Person re-identification (re-ID) aims at identifying the same persons' images across different cameras. However, domain diversities between different datasets pose an evident challenge for adapting the re-ID model trained on one dataset to another one. State-of-the-art unsupervised domain adaptation methods for person ... | MUTUAL MEAN-TEACHING: PSEUDO LABEL REFINERY FOR UNSUPERVISED DO- MAIN ADAPTATION ON PERSON RE-IDENTIFICATION |
d252595735 | Mobile UI understanding is important for enabling various interaction tasks such as UI automation and accessibility. Previous mobile UI modeling often depends on the view hierarchy information of a screen, which directly provides the structural data of the UI, with the hope to bypass challenging tasks of visual modelin... | Published as a conference paper at ICLR 2023 SPOTLIGHT: MOBILE UI UNDERSTANDING USING VISION-LANGUAGE MODELS WITH A FOCUS |
d256231061 | In real-world applications, deep learning models often run in non-stationary environments where the target data distribution continually shifts over time. There have been numerous domain adaptation (DA) methods in both online and offline modes to improve cross-domain adaptation ability. However, these DA methods typica... | Published as a conference paper at ICLR 2023 DEJA VU: CONTINUAL MODEL GENERALIZATION FOR UNSEEN DOMAINS |
d246822597 | The dominant line of work in domain adaptation has focused on learning invariant representations using domain-adversarial training.In this paper, we interpret this approach from a game theoretical perspective.Defining optimal solutions in domain-adversarial training as local Nash equilibria, we show that gradient desce... | DOMAIN ADVERSARIAL TRAINING A GAME PERSPECTIVE |
d246823323 | Diffusion models have emerged as an expressive family of generative models rivaling GANs in sample quality and autoregressive models in likelihood scores. Standard diffusion models typically require hundreds of forward passes through the model to generate a single high-fidelity sample. We introduce Differentiable Diffu... | Published as a conference paper at ICLR 2022 LEARNING FAST SAMPLERS FOR DIFFUSION MODELS BY DIFFERENTIATING THROUGH SAMPLE QUALITY |
d10713737 | We propose a novel training algorithm for reinforcement learning which combines the strength of deep Q-learning with a constrained optimization approach to tighten optimality and encourage faster reward propagation. Our novel technique makes deep reinforcement learning more practical by drastically reducing the trainin... | LEARNING TO PLAY IN A DAY: FASTER DEEP REIN- FORCEMENT LEARNING BY OPTIMALITY TIGHTENING |
d254199184 | Auxiliary tasks improve the representations learned by deep reinforcement learning agents. Analytically, their effect is reasonably well-understood; in practice, however, their primary use remains in support of a main learning objective, rather than as a method for learning representations. This is perhaps surprising g... | Published as a conference paper at ICLR 2023 PROTO-VALUE NETWORKS: SCALING REPRESENTA- TION LEARNING WITH AUXILIARY TASKS |
d249926745 | Mini-batch SGD with momentum is a fundamental algorithm for learning large predictive models. In this paper we develop a new analytic framework to analyze noise-averaged properties of mini-batch SGD for linear models at constant learning rates, momenta and sizes of batches. Our key idea is to consider the dynamics of t... | A VIEW OF MINI-BATCH SGD VIA GENERATING FUNC- TIONS: CONDITIONS OF CONVERGENCE, PHASE TRAN- SITIONS, BENEFIT FROM NEGATIVE MOMENTA |
d10082291 | Despite recent advances, the remaining bottlenecks in deep generative models are necessity of extensive training and difficulties with generalization from small number of training examples. We develop a new generative model called Generative Matching Network which is inspired by the recently proposed matching networks ... | Fast Adaptation in Generative Models with Generative Matching Networks |
d247628080 | Deep metric learning (DML) enables learning with less supervision through its emphasis on the similarity structure of representations. There has been much work on improving generalization of DML in settings like zero-shot retrieval, but little is known about its implications for fairness. In this paper, we are the firs... | IS FAIRNESS ONLY METRIC DEEP? EVALUATING AND ADDRESSING SUBGROUP GAPS IN DML |
d244714783 | DETR is the first end-to-end object detector using a transformer encoder-decoder architecture and demonstrates competitive performance but low computational efficiency on high resolution feature maps. The subsequent work, Deformable DETR, enhances the efficiency of DETR by replacing dense attention with deformable atte... | SPARSE DETR: EFFICIENT END-TO-END OBJECT DE- TECTION WITH LEARNABLE SPARSITY |
d237416749 | Since training a large-scale backdoored model from scratch requires a large training dataset, several recent attacks have considered to inject backdoors into a trained clean model without altering model behaviors on the clean data. Previous work finds that backdoors can be injected into a trained clean model with Adver... | HOW TO INJECT BACKDOORS WITH BETTER CONSIS- TENCY: LOGIT ANCHORING ON CLEAN DATA |
d159268776 | regional laws in a comparative perspective.The selection of the two laws is based on the preliminary study which found a quite unique form of criminal provisions on each laws.The analysis is also based on art 200 and 201 Law No. 36/2009 and its derivative regulations as a normative measurement in national level, with w... | Comparative Study on Criminal Provisions on Regional Regulations Concerning Exclusive Breastfeeding |
d3463260 | We propose a distributed architecture for deep reinforcement learning at scale, that enables agents to learn effectively from orders of magnitude more data than previously possible. The algorithm decouples acting from learning: the actors interact with their own instances of the environment by selecting actions accordi... | Published as a conference paper at ICLR 2018 DISTRIBUTED PRIORITIZED EXPERIENCE REPLAY |
d13805769 | Distributed representations of meaning are a natural way to encode covariance relationships between words and phrases in NLP. By overcoming data sparsity problems, as well as providing information about semantic relatedness which is not available in discrete representations, distributed representations have proven usef... | Multilingual Distributed Representations without Word Alignment |
d213729382 | Deep neural networks often have millions of parameters. This can hinder their deployment to low-end devices, not only due to high memory requirements but also because of increased latency at inference. We propose a novel model compression method that generates a sparse trained model without additional overhead: by allo... | Published as a conference paper at ICLR 2020 DYNAMIC MODEL PRUNING WITH FEEDBACK |
d174801567 | The lottery ticket hypothesis proposes that over-parameterization of deep neural networks (DNNs) aids training by increasing the probability of a "lucky" sub-network initialization being present rather than by helping the optimization process (Frankle & Carbin, 2019). Intriguingly, this phenomenon suggests that initial... | Published as a conference paper at ICLR 2020 PLAYING THE LOTTERY WITH REWARDS AND MULTIPLE LANGUAGES: LOTTERY TICKETS IN RL AND NLP |
d8404331 | Deep neural networks are easily fooled: High confidence predictions for unrecognizable images. In | Workshop track -ICLR 2017 ROBUSTNESS TO ADVERSARIAL EXAMPLES THROUGH AN ENSEMBLE OF SPECIALISTS |
d1107124 | We propose a reparameterization of LSTM that brings the benefits of batch normalization to recurrent neural networks. Whereas previous works only apply batch normalization to the input-to-hidden transformation of RNNs, we demonstrate that it is both possible and beneficial to batch-normalize the hidden-to-hidden transi... | Published as a conference paper at ICLR 2017 RECURRENT BATCH NORMALIZATION |
d209532006 | We propose and address a novel few-shot RL problem, where a task is characterized by a subtask graph which describes a set of subtasks and their dependencies that are unknown to the agent. The agent needs to quickly adapt to the task over few episodes during adaptation phase to maximize the return in the test phase. In... | Published as a conference paper at ICLR 2020 META REINFORCEMENT LEARNING WITH AUTONOMOUS INFERENCE OF SUBTASK DEPENDENCIES |
d212877887 | The success of reinforcement learning for real world robotics has been, in many cases limited to instrumented laboratory scenarios, often requiring arduous human effort and oversight to enable continuous learning. In this work, we discuss the elements that are needed for a robotic learning system that can continually a... | THE INGREDIENTS OF REAL-WORLD ROBOTIC REINFORCEMENT LEARNING |
d246411523 | The implicit bias induced by the training of neural networks has become a topic of rigorous study. In the limit of gradient flow and gradient descent with appropriate step size, it has been shown that when one trains a deep linear network with logistic or exponential loss on linearly separable data, the weights converg... | Published as a conference paper at ICLR 2022 TRAINING INVARIANCES AND THE LOW-RANK PHE- NOMENON: BEYOND LINEAR NETWORKS |
d235391000 | While large-scale pretrained language models have obtained impressive results when fine-tuned on a wide variety of tasks, they still often suffer from overfitting in low-resource scenarios. Since such models are general-purpose feature extractors, many of these features are inevitably irrelevant for a given target task... | Published as a conference paper at ICLR 2021 VARIATIONAL INFORMATION BOTTLENECK FOR EFFEC- TIVE LOW-RESOURCE FINE-TUNING |
d254275077 | The ability to quickly and accurately identify covariate shift at test time is a critical and often overlooked component of safe machine learning systems deployed in high-risk domains.While methods exist for detecting when predictions should not be made on out-of-distribution test examples, identifying distributional l... | A LEARNING BASED HYPOTHESIS TEST FOR HARM-FUL COVARIATE SHIFT |
d211133181 | Generative models are often used to sample high-dimensional data points from a manifold with small intrinsic dimension. Existing techniques for comparing generative models focus on global data properties such as mean and covariance; in that sense, they are extrinsic and uni-scale. We develop the first, to our knowledge... | Intrinsic Multi-scale Evaluation of Generative Models |
d219966188 | We analyze the convergence of the averaged stochastic gradient descent for overparameterized two-layer neural networks for regression problems. It was recently found that a neural tangent kernel (NTK) plays an important role in showing the global convergence of gradient-based methods under the NTK regime, where the lea... | OPTIMAL RATES FOR AVERAGED STOCHASTIC GRA- DIENT DESCENT UNDER NEURAL TANGENT KERNEL REGIME |
d238407772 | We consider the problem of training a classification model with group annotated training data. Recent work has established that, if there is distribution shift across different groups, models trained using the standard empirical risk minimization (ERM) objective suffer from poor performance on minority groups and that ... | Published as a conference paper at ICLR 2022 FOCUS ON THE COMMON GOOD: GROUP DISTRIBU- TIONAL ROBUSTNESS FOLLOWS |
d3699386 | Despite being impactful on a variety of problems and applications, the generative adversarial nets (GANs) are remarkably difficult to train. This issue is formally analyzed by , who also propose an alternative direction to avoid the caveats in the minmax two-player training of GANs. The corresponding algorithm, called ... | Published as a conference paper at ICLR 2018 IMPROVING THE IMPROVED TRAINING OF WASSERSTEIN GANS: A CONSISTENCY TERM AND ITS DUAL EFFECT |
d211146532 | Overparameterization has been shown to benefit both the optimization and generalization of neural networks, but large networks are resource hungry at both training and test time. Network pruning can reduce test-time resource requirements, but is typically applied to trained networks and therefore cannot avoid the expen... | Published as a conference paper at ICLR 2020 PICKING WINNING TICKETS BEFORE TRAINING BY PRESERVING GRADIENT FLOW |
d68221207 | Partial differential equations (PDEs) are widely used across the physical and computational sciences. Decades of research and engineering went into designing fast iterative solution methods. Existing solvers are general purpose, but may be sub-optimal for specific classes of problems. In contrast to existing hand-craft... | LEARNING NEURAL PDE SOLVERS WITH CONVER- GENCE GUARANTEES |
d256105154 | In computer vision, it is often observed that formulating regression problems as a classification task yields better performance. We investigate this curious phenomenon and provide a derivation to show that classification, with the crossentropy loss, outperforms regression with a mean squared error loss in its ability ... | Published as a conference paper at ICLR 2023 IMPROVING DEEP REGRESSION WITH ORDINAL EN- TROPY |
d257232381 | Deep neural networks are likely to fail when the test data is corrupted in realworld deployment (e.g., blur, weather, etc.). Test-time optimization is an effective way that adapts models to generalize to corrupted data during testing, which has been shown in the image domain. However, the techniques for improving video... | Published as a conference paper at ICLR 2023 TEMPORAL COHERENT TEST-TIME OPTIMIZATION FOR ROBUST VIDEO CLASSIFICATION |
d232269775 | Distributionally robust optimization (DRO) provides a framework for training machine learning models that are able to perform well on a collection of related data distributions (the "uncertainty set"). This is done by solving a min-max game: the model is trained to minimize its maximum expected loss among all distribut... | Published as a conference paper at ICLR 2021 MODELING THE SECOND PLAYER IN DISTRIBUTIONALLY ROBUST OPTIMIZATION |
d252968162 | We consider Contextual Bandits with Concave Rewards (CBCR), a multi-objective bandit problem where the desired trade-off between the rewards is defined by a known concave objective function, and the reward vector depends on an observed stochastic context. We present the first algorithm with provably vanishing regret fo... | Published as a conference paper at ICLR 2023 CONTEXTUAL BANDITS WITH CONCAVE REWARDS, AND AN APPLICATION TO FAIR RANKING |
d252780995 | To obtain, deterministic guarantees of adversarial robustness, specialized training methods are used. We propose, SABR, a novel such certified training method, based on the key insight that propagating interval bounds for a small but carefully selected subset of the adversarial input region is sufficient to approximate... | Published as a conference paper at ICLR 2023 CERTIFIED TRAINING: SMALL BOXES ARE ALL YOU NEED |
d238419059 | In contrast to single-objective optimization (SOO), multi-objective optimization (MOO) requires an optimizer to find the Pareto frontier, a subset of feasible solutions that are not dominated by other feasible solutions. In this paper, we propose LaMOO, a novel multi-objective optimizer that learns a model from observe... | MULTI-OBJECTIVE OPTIMIZATION BY LEARNING SPACE PARTITIONS |
d257365443 | In Natural Language Processing (NLP), intelligent neuron models can be susceptible to textual Trojan attacks. Such attacks occur when Trojan models behave normally for standard inputs but generate malicious output for inputs that contain a specific trigger. Syntactic-structure triggers, which are invisible, are becomin... | |
d231603061 | We study the multi-agent safe control problem where agents should avoid collisions to static obstacles and collisions with each other while reaching their goals. Our core idea is to learn the multi-agent control policy jointly with learning the control barrier functions as safety certificates. We propose a novel joint-... | Published as a conference paper at ICLR 2021 LEARNING SAFE MULTI-AGENT CONTROL WITH DECENTRALIZED NEURAL BARRIER CERTIFICATES |
d234334847 | Well-designed molecular representations (fingerprints) are vital to combine medical chemistry and deep learning. Whereas incorporating 3D geometry of molecules (i.e. conformations) in their representations seems beneficial, current 3D algorithms are still in infancy. In this paper, we propose a novel molecular represen... | Published as a conference paper at ICLR 2021 HAMNET: CONFORMATION-GUIDED MOLECULAR REPRESENTATION WITH HAMILTONIAN NEURAL NET- WORKS |
d10316648 | In this work, we present a novel neural network based architecture for inducing compositional crosslingual word representations. Unlike previously proposed methods, our method fulfills the following three criteria; it constrains the wordlevel representations to be compositional, it is capable of leveraging both bilingu... | LEVERAGING MONOLINGUAL DATA FOR CROSSLIN- GUAL COMPOSITIONAL WORD REPRESENTATIONS |
d210838871 | As deep neural networks (DNNs) achieve tremendous success across many application domains, researchers tried to explore in many aspects on why they generalize well. In this paper, we provide a novel perspective on these issues using the gradient signal to noise ratio (GSNR) of parameters during training process of DNNs... | Published as a conference paper at ICLR 2020 UNDERSTANDING WHY NEURAL NETWORKS GENER- ALIZE WELL THROUGH GSNR OF PARAMETERS |
d59523607 | We introduce a framework for Continual Learning (CL) based on Bayesian inference over the function space rather than the parameters of a deep neural network. This method, referred to as functional regularisation for Continual Learning, avoids forgetting a previous task by constructing and memorising an approximate post... | Published as a conference paper at ICLR 2020 FUNCTIONAL REGULARISATION FOR CONTINUAL LEARNING WITH GAUSSIAN PROCESSES |
d210845646 | We propose to study the problem of few-shot graph classification in graph neural networks (GNNs) to recognize unseen classes, given limited labeled graph examples. Despite several interesting GNN variants being proposed recently for node and graph classification tasks, when faced with scarce labeled examples in the few... | FEW-SHOT LEARNING ON GRAPHS VIA SUPER- CLASSES BASED ON GRAPH SPECTRAL MEASURES |
d252682915 | Accurate delineation of fine-scale structures is a very important yet challenging problem. Existing methods use topological information as an additional training loss, but are ultimately making pixel-wise predictions. In this paper, we propose the first deep learning based method to learn topological/structural represe... | LEARNING PROBABILISTIC TOPOLOGICAL REPRESEN- TATIONS USING DISCRETE MORSE THEORY |
d219531522 | Non-autoregressive text to speech (TTS) models such as FastSpeech(Ren et al., 2019)can synthesize speech significantly faster than previous autoregressive models with comparable quality. The training of FastSpeech model relies on an autoregressive teacher model for duration prediction (to provide more information as in... | FASTSPEECH 2: FAST AND HIGH-QUALITY END-TO- END TEXT TO SPEECH |
d226964694 | Top-k predictions are used in many real-world applications such as machine learning as a service, recommender systems, and web searches. 0 -norm adversarial perturbation characterizes an attack that arbitrarily modifies some features of an input such that a classifier makes an incorrect prediction for the perturbed inp... | Published as a conference paper at ICLR 2022 ALMOST TIGHT L0-NORM CERTIFIED ROBUSTNESS OF TOP-k PREDICTIONS AGAINST ADVERSARIAL PERTUR- BATIONS |
d10766488 | In this paper, we introduce a novel deep learning framework, termed Purine. In Purine, a deep network is expressed as a bipartite graph (bi-graph), which is composed of interconnected operators and data tensors. With the bi-graph abstraction, networks are easily solvable with event-driven task dispatcher. We then demon... | PURINE: A BI-GRAPH BASED DEEP LEARNING FRAME- WORK |
d254044710 | Recent improvements in conditional generative modeling have made it possible to generate high-quality images from language descriptions alone. We investigate whether these methods can directly address the problem of sequential decisionmaking. We view decision-making not through the lens of reinforcement learning (RL), ... | IS CONDITIONAL GENERATIVE MODELING ALL YOU NEED FOR DECISION-MAKING? |
d248239874 | We consider the offline constrained reinforcement learning (RL) problem, in which the agent aims to compute a policy that maximizes expected return while satisfying given cost constraints, learning only from a pre-collected dataset. This problem setting is appealing in many real-world scenarios, where direct interactio... | COPTIDICE: OFFLINE CONSTRAINED REINFORCE- MENT LEARNING VIA STATIONARY DISTRIBUTION CORRECTION ESTIMATION |
d256503890 | De novo molecular generation is an essential task for science discovery. Recently, fragment-based deep generative models have attracted much research attention due to their flexibility in generating novel molecules based on existing molecule fragments. However, the motif vocabulary, i.e., the collection of frequent fra... | Published as a conference paper at ICLR 2023 DE NOVO MOLECULAR GENERATION VIA CONNECTION-AWARE MOTIF MINING |
d250408169 | Transformers have quickly shined in the computer vision world since the emergence of Vision Transformers (ViTs). The dominant role of convolutional neural networks (CNNs) seems to be challenged by increasingly effective transformer-based models. Very recently, a couple of advanced convolutional models strike back with ... | MORE CONVNETS IN THE 2020S: SCALING UP KER- NELS BEYOND 51 × 51 USING SPARSITY |
d211171395 | Inductive representation learning on temporal graphs is an important step toward salable machine learning on real-world dynamic networks. The evolving nature of temporal dynamic graphs requires handling new nodes as well as capturing temporal patterns. The node embeddings, which are now functions of time, should repres... | Published as a conference paper at ICLR 2020 INDUCTIVE REPRESENTATION LEARNING ON TEMPORAL GRAPHS |
d231786434 | Zero-shot learning (ZSL) aims to classify images of an unseen class only based on a few attributes describing that class but no access to any training sample. A popular strategy is to learn a mapping between the semantic space of class attributes and the visual space of images based on the seen classes and their data. ... | Published as a conference paper at ICLR 2021 ISOMETRIC PROPAGATION NETWORK FOR GENERALIZED ZERO-SHOT LEARNING |
d3286670 | Deep neural networks are surprisingly efficient at solving practical tasks, but the theory behind this phenomenon is only starting to catch up with the practice. Numerous works show that depth is the key to this efficiency. A certain class of deep convolutional networks -namely those that correspond to the Hierarchical... | EXPRESSIVE POWER OF RECURRENT NEURAL NET- WORKS |
d259108495 | Recently, sequence learning methods have been applied to the problem of off-policy Reinforcement Learning, including the seminal work on Decision Transformers, which employs transformers for this task. Since transformers are parameter-heavy, cannot benefit from history longer than a fixed window size, and are not compu... | Published as a conference paper at ICLR 2023 DECISION S4: EFFICIENT SEQUENCE-BASED RL VIA STATE SPACE LAYERS |
d234501004 | Academic trade requires juggling multiple variants of the same content published in different formats: manuscripts, presentations, posters and computational notebooks. The need to track versions to accommodate for the write-review-rebutrevise life-cycle adds another layer of complexity. We propose to significantly redu... | Published at Rethinking ML Papers -ICLR 2021 Workshop YOU ONLY WRITE THRICE: CREATING DOCUMENTS, COMPUTATIONAL NOTE- BOOKS AND PRESENTATIONS FROM A SINGLE SOURCE |
d16053260 | Learning compact, interpretable image representations is a very natural task which has not been solved satisfactorily even for simple classes of binary images. In this paper, we review various ways of composing parts (or experts) for binary data and argue that competitive forms of interaction are best suited to learn l... | COMPACT PART-BASED IMAGE REPRESENTATIONS |
d253080440 | This paper focuses on computing the convex conjugate operation that arises when solving Euclidean Wasserstein-2 optimal transport problems. This conjugation, which is also referred to as the Legendre-Fenchel conjugate or c-transform, is considered difficult to compute and in practice, Wasserstein-2 methods are limited ... | Published as a conference paper at ICLR 2023 ON AMORTIZING CONVEX CONJUGATES FOR OPTIMAL TRANSPORT |
d249605400 | While machine learning models rapidly advance the state-of-the-art on various real-world tasks, out-of-domain (OOD) generalization remains a challenging problem given the vulnerability of these models to spurious correlations. We propose a balanced mini-batch sampling strategy to transform a biased data distribution in... | Published as a conference paper at ICLR 2023 CAUSAL BALANCING FOR DOMAIN GENERALIZATION |
d249191434 | The problem of optimization on Stiefel manifold, i.e., minimizing functions of (not necessarily square) matrices that satisfy orthogonality constraints, has been extensively studied. Yet, a new approach is proposed based on, for the first time, an interplay between thoughtfully designed continuous and discrete dynamics... | MOMENTUM STIEFEL OPTIMIZER, WITH APPLICA- TIONS TO SUITABLY-ORTHOGONAL ATTENTION, AND OPTIMAL TRANSPORT |
d256826892 | In deep learning, transferring information from a pretrained network to a downstream task by finetuning has many benefits. The choice of task head plays an important role in fine-tuning, as the pretrained and downstream tasks are usually different. Although there exist many different designs for finetuning, a full unde... | Published as a conference paper at ICLR 2023 HOW TO PREPARE YOUR TASK HEAD FOR FINETUNING |
d226282381 | We show when maximizing a properly defined f -divergence measure with respect to a classifier's predictions and the supervised labels is robust with label noise. Leveraging its variational form, we derive a nice decoupling property for a family of f -divergence measures when label noise presents, where the divergence i... | Published as a conference paper at ICLR 2021 WHEN OPTIMIZING f -DIVERGENCE IS ROBUST WITH LABEL NOISE |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.