title
stringlengths
12
151
url
stringlengths
41
43
detail_url
stringlengths
41
43
authors
stringlengths
6
562
tags
stringclasses
3 values
abstract
stringlengths
519
2.34k
pdf
stringlengths
71
71
Learning Towards The Largest Margins
https://openreview.net/forum?id=hqkhcFHOeKD
https://openreview.net/forum?id=hqkhcFHOeKD
Xiong Zhou,Xianming Liu,Deming Zhai,Junjun Jiang,Xin Gao,Xiangyang Ji
ICLR 2022,Poster
One of the main challenges for feature representation in deep learning-based classification is the design of appropriate loss functions that exhibit strong discriminative power. The classical softmax loss does not explicitly encourage discriminative learning of features. A popular direction of research is to incorporat...
https://openreview.net/pdf/05f12453b1762c08d54507567f592f91d86425be.pdf
Patch-Fool: Are Vision Transformers Always Robust Against Adversarial Perturbations?
https://openreview.net/forum?id=28ib9tf6zhr
https://openreview.net/forum?id=28ib9tf6zhr
Yonggan Fu,Shunyao Zhang,Shang Wu,Cheng Wan,Yingyan Lin
ICLR 2022,Poster
Vision transformers (ViTs) have recently set off a new wave in neural architecture design thanks to their record-breaking performance in various vision tasks. In parallel, to fulfill the goal of deploying ViTs into real-world vision applications, their robustness against potential malicious attacks has gained increasin...
https://openreview.net/pdf/4c7b8d2f80c4ea1bfe11754da2e7c69fc5183754.pdf
AdaMatch: A Unified Approach to Semi-Supervised Learning and Domain Adaptation
https://openreview.net/forum?id=Q5uh1Nvv5dm
https://openreview.net/forum?id=Q5uh1Nvv5dm
David Berthelot,Rebecca Roelofs,Kihyuk Sohn,Nicholas Carlini,Alexey Kurakin
ICLR 2022,Poster
We extend semi-supervised learning to the problem of domain adaptation to learn significantly higher-accuracy models that train on one data distribution and test on a different one. With the goal of generality, we introduce AdaMatch, a unified solution for unsupervised domain adaptation (UDA), semi-supervised learning ...
https://openreview.net/pdf/8dd30c7eff2e4f152d2d24368c232baec4e5e974.pdf
Complete Verification via Multi-Neuron Relaxation Guided Branch-and-Bound
https://openreview.net/forum?id=l_amHf1oaK
https://openreview.net/forum?id=l_amHf1oaK
Claudio Ferrari,Mark Niklas Mueller,Nikola Jovanović,Martin Vechev
ICLR 2022,Poster
State-of-the-art neural network verifiers are fundamentally based on one of two paradigms: either encoding the whole verification problem via tight multi-neuron convex relaxations or applying a Branch-and-Bound (BaB) procedure leveraging imprecise but fast bounding methods on a large number of easier subproblems. The f...
https://openreview.net/pdf/fcc20218f5754386cf64f4156a1f41039038b5da.pdf
Learning Fast Samplers for Diffusion Models by Differentiating Through Sample Quality
https://openreview.net/forum?id=VFBjuF8HEp
https://openreview.net/forum?id=VFBjuF8HEp
Daniel Watson,William Chan,Jonathan Ho,Mohammad Norouzi
ICLR 2022,Poster
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...
https://openreview.net/pdf/56f0145dd15f32bd53f6dba7efde74914a88f663.pdf
Distribution Compression in Near-Linear Time
https://openreview.net/forum?id=lzupY5zjaU9
https://openreview.net/forum?id=lzupY5zjaU9
Abhishek Shetty,Raaz Dwivedi,Lester Mackey
ICLR 2022,Poster
In distribution compression, one aims to accurately summarize a probability distribution $\mathbb{P}$ using a small number of representative points. Near-optimal thinning procedures achieve this goal by sampling $n$ points from a Markov chain and identifying $\sqrt{n}$ points with $\widetilde{\mathcal{O}}(1/\sqrt{n})$ ...
https://openreview.net/pdf/484f68f97f561be1f3272522336a9a0b1fa84bbc.pdf
Capturing Structural Locality in Non-parametric Language Models
https://openreview.net/forum?id=nnU3IUMJmN
https://openreview.net/forum?id=nnU3IUMJmN
Frank F. Xu,Junxian He,Graham Neubig,Vincent Josua Hellendoorn
ICLR 2022,Poster
Structural locality is a ubiquitous feature of real-world datasets, wherein data points are organized into local hierarchies. Some examples include topical clusters in text or project hierarchies in source code repositories. In this paper, we explore utilizing this structural locality within non-parametric language mod...
https://openreview.net/pdf/05677eb0d7fca88dd7c4c6cbefa73f6ae430ad68.pdf
Audio Lottery: Speech Recognition Made Ultra-Lightweight, Noise-Robust, and Transferable
https://openreview.net/forum?id=9Nk6AJkVYB
https://openreview.net/forum?id=9Nk6AJkVYB
Shaojin Ding,Tianlong Chen,Zhangyang Wang
ICLR 2022,Poster
Lightweight speech recognition models have seen explosive demands owing to a growing amount of speech-interactive features on mobile devices. Since designing such systems from scratch is non-trivial, practitioners typically choose to compress large (pre-trained) speech models. Recently, lottery ticket hypothesis reveal...
https://openreview.net/pdf/3d42ff881f8ec8954935d0f8bbcb2a21d71106ea.pdf
Learning to Map for Active Semantic Goal Navigation
https://openreview.net/forum?id=swrMQttr6wN
https://openreview.net/forum?id=swrMQttr6wN
Georgios Georgakis,Bernadette Bucher,Karl Schmeckpeper,Siddharth Singh,Kostas Daniilidis
ICLR 2022,Poster
We consider the problem of object goal navigation in unseen environments. Solving this problem requires learning of contextual semantic priors, a challenging endeavour given the spatial and semantic variability of indoor environments. Current methods learn to implicitly encode these priors through goal-oriented navigat...
https://openreview.net/pdf/8097afd8a3e6d7c824f59390ca5a9cee0530bbd1.pdf
Benchmarking the Spectrum of Agent Capabilities
https://openreview.net/forum?id=1W0z96MFEoH
https://openreview.net/forum?id=1W0z96MFEoH
Danijar Hafner
ICLR 2022,Poster
Evaluating the general abilities of intelligent agents requires complex simulation environments. Existing benchmarks typically evaluate only one narrow task per environment, requiring researchers to perform expensive training runs on many different environments. We introduce Crafter, an open world survival game with vi...
https://openreview.net/pdf/116a18888b3fb460e882ec2b844128223e3b17ca.pdf
Mind the Gap: Domain Gap Control for Single Shot Domain Adaptation for Generative Adversarial Networks
https://openreview.net/forum?id=vqGi8Kp0wM
https://openreview.net/forum?id=vqGi8Kp0wM
Peihao Zhu,Rameen Abdal,John Femiani,Peter Wonka
ICLR 2022,Poster
We present a new method for one shot domain adaptation. The input to our method is trained GAN that can produce images in domain A and a single reference image I_B from domain B. The proposed algorithm can translate any output of the trained GAN from domain A to domain B. There are two main advantages of our method com...
https://openreview.net/pdf/2f6e593f100fa850ecde50e059aa6b2e73a3f6fe.pdf
On Evaluation Metrics for Graph Generative Models
https://openreview.net/forum?id=EnwCZixjSh
https://openreview.net/forum?id=EnwCZixjSh
Rylee Thompson,Boris Knyazev,Elahe Ghalebi,Jungtaek Kim,Graham W. Taylor
ICLR 2022,Poster
In image generation, generative models can be evaluated naturally by visually inspecting model outputs. However, this is not always the case for graph generative models (GGMs), making their evaluation challenging. Currently, the standard process for evaluating GGMs suffers from three critical limitations: i) it does no...
https://openreview.net/pdf/fcb94055fd54a7db263aab7d0f85b591c34e713e.pdf
Selective Ensembles for Consistent Predictions
https://openreview.net/forum?id=HfUyCRBeQc
https://openreview.net/forum?id=HfUyCRBeQc
Emily Black,Klas Leino,Matt Fredrikson
ICLR 2022,Poster
Recent work has shown that models trained to the same objective, and which achieve similar measures of accuracy on consistent test data, may nonetheless behave very differently on individual predictions. This inconsistency is undesirable in high-stakes contexts, such as medical diagnosis and finance. We show that this ...
https://openreview.net/pdf/aef96c65d43466af59147df0d990f0b94efbef7a.pdf
Graph Condensation for Graph Neural Networks
https://openreview.net/forum?id=WLEx3Jo4QaB
https://openreview.net/forum?id=WLEx3Jo4QaB
Wei Jin,Lingxiao Zhao,Shichang Zhang,Yozen Liu,Jiliang Tang,Neil Shah
ICLR 2022,Poster
Given the prevalence of large-scale graphs in real-world applications, the storage and time for training neural models have raised increasing concerns. To alleviate the concerns, we propose and study the problem of graph condensation for graph neural networks (GNNs). Specifically, we aim to condense the large, origina...
https://openreview.net/pdf/fb904d1d840eb264e6ab2e160ff7322153a1fbb0.pdf
DIVA: Dataset Derivative of a Learning Task
https://openreview.net/forum?id=bVvMOtLMiw
https://openreview.net/forum?id=bVvMOtLMiw
Yonatan Dukler,Alessandro Achille,Giovanni Paolini,Avinash Ravichandran,Marzia Polito,Stefano Soatto
ICLR 2022,Poster
We present a method to compute the derivative of a learning task with respect to a dataset. A learning task is a function from a training set to the validation error, which can be represented by a trained deep neural network (DNN). The ``dataset derivative'' is a linear operator, computed around the trained model, that...
https://openreview.net/pdf/c20ae574c689fe5fbecb96f791b3e678973e0053.pdf
Towards General Function Approximation in Zero-Sum Markov Games
https://openreview.net/forum?id=sA4qIu3zv6v
https://openreview.net/forum?id=sA4qIu3zv6v
Baihe Huang,Jason D. Lee,Zhaoran Wang,Zhuoran Yang
ICLR 2022,Poster
This paper considers two-player zero-sum finite-horizon Markov games with simultaneous moves. The study focuses on the challenging settings where the value function or the model is parameterized by general function classes. Provably efficient algorithms for both decoupled and coordinated settings are developed. In the ...
https://openreview.net/pdf/89164a5698b4ced1396254451108620fc52d5bc1.pdf
Exposing the Implicit Energy Networks behind Masked Language Models via Metropolis--Hastings
https://openreview.net/forum?id=6PvWo1kEvlT
https://openreview.net/forum?id=6PvWo1kEvlT
Kartik Goyal,Chris Dyer,Taylor Berg-Kirkpatrick
ICLR 2022,Poster
While recent work has shown that scores from models trained by the ubiquitous masked language modeling (MLM) objective effectively discriminate probable from improbable sequences, it is still an open question if these MLMs specify a principled probability distribution over the space of possible sequences. In this paper...
https://openreview.net/pdf/dfdc7212f0c035baaec71e0d9d64317aec15492b.pdf
ClimateGAN: Raising Climate Change Awareness by Generating Images of Floods
https://openreview.net/forum?id=EZNOb_uNpJk
https://openreview.net/forum?id=EZNOb_uNpJk
Victor Schmidt,Alexandra Luccioni,Mélisande Teng,Tianyu Zhang,Alexia Reynaud,Sunand Raghupathi,Gautier Cosne,Adrien Juraver,Vahe Vardanyan,Alex Hernández-García,Yoshua Bengio
ICLR 2022,Poster
Climate change is a major threat to humanity and the actions required to prevent its catastrophic consequences include changes in both policy-making and individual behaviour. However, taking action requires understanding its seemingly abstract and distant consequences. Projecting the potential impacts of extreme climat...
https://openreview.net/pdf/ca121d72177c0fb77244bde0b2958681a89d4b98.pdf
A Comparison of Hamming Errors of Representative Variable Selection Methods
https://openreview.net/forum?id=nhN-fqxmNGx
https://openreview.net/forum?id=nhN-fqxmNGx
Tracy Ke,Longlin Wang
ICLR 2022,Poster
Lasso is a celebrated method for variable selection in linear models, but it faces challenges when the covariates are moderately or strongly correlated. This motivates alternative approaches such as using a non-convex penalty, adding a ridge regularization, or conducting a post-Lasso thresholding. In this paper, we com...
https://openreview.net/pdf/ae8e44624ed225194ef2c6ef294ae6d5067515b8.pdf
A Program to Build E(N)-Equivariant Steerable CNNs
https://openreview.net/forum?id=WE4qe9xlnQw
https://openreview.net/forum?id=WE4qe9xlnQw
Gabriele Cesa,Leon Lang,Maurice Weiler
ICLR 2022,Poster
Equivariance is becoming an increasingly popular design choice to build data efficient neural networks by exploiting prior knowledge about the symmetries of the problem at hand. Euclidean steerable CNNs are one of the most common classes of equivariant networks. While the constraints these architectures need to satisfy...
https://openreview.net/pdf/6d634b6f1eabc70593f897e223c78025e3029b52.pdf
Minimax Optimization with Smooth Algorithmic Adversaries
https://openreview.net/forum?id=UdxJ2fJx7N0
https://openreview.net/forum?id=UdxJ2fJx7N0
Tanner Fiez,Chi Jin,Praneeth Netrapalli,Lillian J Ratliff
ICLR 2022,Poster
This paper considers minimax optimization $\min_x \max_y f(x, y)$ in the challenging setting where $f$ can be both nonconvex in $x$ and nonconcave in $y$. Though such optimization problems arise in many machine learning paradigms including training generative adversarial networks (GANs) and adversarially robust models,...
https://openreview.net/pdf/6f978c34600cf6fcf440c6e1bf8d1f93e0afce3d.pdf
On Distributed Adaptive Optimization with Gradient Compression
https://openreview.net/forum?id=CI-xXX9dg9l
https://openreview.net/forum?id=CI-xXX9dg9l
Xiaoyun Li,Belhal Karimi,Ping Li
ICLR 2022,Poster
We study COMP-AMS, a distributed optimization framework based on gradient averaging and adaptive AMSGrad algorithm. Gradient compression with error feedback is applied to reduce the communication cost in the gradient transmission process. Our convergence analysis of COMP-AMS shows that such compressed gradient averagin...
https://openreview.net/pdf/84313c8e0bf7b65d71addc3b16aba48f161f4092.pdf
Leveraging unlabeled data to predict out-of-distribution performance
https://openreview.net/forum?id=o_HsiMPYh_x
https://openreview.net/forum?id=o_HsiMPYh_x
Saurabh Garg,Sivaraman Balakrishnan,Zachary Chase Lipton,Behnam Neyshabur,Hanie Sedghi
ICLR 2022,Poster
Real-world machine learning deployments are characterized by mismatches between the source (training) and target (test) distributions that may cause performance drops. In this work, we investigate methods for predicting the target domain accuracy using only labeled source data and unlabeled target data. We propose Aver...
https://openreview.net/pdf/f94008d1c0cfc4177d8617db211b62b1f85906ea.pdf
VC dimension of partially quantized neural networks in the overparametrized regime
https://openreview.net/forum?id=7udZAsEzd60
https://openreview.net/forum?id=7udZAsEzd60
Yutong Wang,Clayton Scott
ICLR 2022,Poster
Vapnik-Chervonenkis (VC) theory has so far been unable to explain the small generalization error of overparametrized neural networks. Indeed, existing applications of VC theory to large networks obtain upper bounds on VC dimension that are proportional to the number of weights, and for a large class of networks, these ...
https://openreview.net/pdf/9760187606b3496a5f4a0fe752a22416bb4a2e21.pdf
Optimal Representations for Covariate Shift
https://openreview.net/forum?id=Rf58LPCwJj0
https://openreview.net/forum?id=Rf58LPCwJj0
Yangjun Ruan,Yann Dubois,Chris J. Maddison
ICLR 2022,Poster
Machine learning systems often experience a distribution shift between training and testing. In this paper, we introduce a simple variational objective whose optima are exactly the set of all representations on which risk minimizers are guaranteed to be robust to any distribution shift that preserves the Bayes predicto...
https://openreview.net/pdf/ddc6369b11aed2bc1a72bc2f493bb2ebd0f65be7.pdf
Fortuitous Forgetting in Connectionist Networks
https://openreview.net/forum?id=ei3SY1_zYsE
https://openreview.net/forum?id=ei3SY1_zYsE
Hattie Zhou,Ankit Vani,Hugo Larochelle,Aaron Courville
ICLR 2022,Poster
Forgetting is often seen as an unwanted characteristic in both human and machine learning. However, we propose that forgetting can in fact be favorable to learning. We introduce forget-and-relearn as a powerful paradigm for shaping the learning trajectories of artificial neural networks. In this process, the forgetting...
https://openreview.net/pdf/ca4d5fd0fac40867b797ca356f4056c7cb11fc6a.pdf
EigenGame Unloaded: When playing games is better than optimizing
https://openreview.net/forum?id=So6YAqnqgMj
https://openreview.net/forum?id=So6YAqnqgMj
Ian Gemp,Brian McWilliams,Claire Vernade,Thore Graepel
ICLR 2022,Poster
We build on the recently proposed EigenGame that views eigendecomposition as a competitive game. EigenGame's updates are biased if computed using minibatches of data, which hinders convergence and more sophisticated parallelism in the stochastic setting. In this work, we propose an unbiased stochastic update that is as...
https://openreview.net/pdf/cedcb096f43d8f1b1e43c8969cf5b1dd7e83d5ae.pdf
Contextualized Scene Imagination for Generative Commonsense Reasoning
https://openreview.net/forum?id=Oh1r2wApbPv
https://openreview.net/forum?id=Oh1r2wApbPv
PeiFeng Wang,Jonathan Zamora,Junfeng Liu,Filip Ilievski,Muhao Chen,Xiang Ren
ICLR 2022,Poster
Humans use natural language to compose common concepts from their environment into plausible, day-to-day scene descriptions. However, such generative commonsense reasoning (GCSR) skills are lacking in state-of-the-art text generation methods. Descriptive sentences about arbitrary concepts generated by neural text gener...
https://openreview.net/pdf/a66e1b12b2211131a44463611c8c272c21decbfb.pdf
Scene Transformer: A unified architecture for predicting future trajectories of multiple agents
https://openreview.net/forum?id=Wm3EA5OlHsG
https://openreview.net/forum?id=Wm3EA5OlHsG
Jiquan Ngiam,Vijay Vasudevan,Benjamin Caine,Zhengdong Zhang,Hao-Tien Lewis Chiang,Jeffrey Ling,Rebecca Roelofs,Alex Bewley,Chenxi Liu,Ashish Venugopal,David J Weiss,Benjamin Sapp,Zhifeng Chen,Jonathon Shlens
ICLR 2022,Poster
Predicting the motion of multiple agents is necessary for planning in dynamic environments. This task is challenging for autonomous driving since agents (e.g., vehicles and pedestrians) and their associated behaviors may be diverse and influence one another. Most prior work have focused on predicting independent future...
https://openreview.net/pdf/92f191f2cdcf1389ed2d3dce901833dc5fc6deaf.pdf
DISSECT: Disentangled Simultaneous Explanations via Concept Traversals
https://openreview.net/forum?id=qY79G8jGsep
https://openreview.net/forum?id=qY79G8jGsep
Asma Ghandeharioun,Been Kim,Chun-Liang Li,Brendan Jou,Brian Eoff,Rosalind Picard
ICLR 2022,Poster
Explaining deep learning model inferences is a promising venue for scientific understanding, improving safety, uncovering hidden biases, evaluating fairness, and beyond, as argued by many scholars. One of the principal benefits of counterfactual explanations is allowing users to explore "what-if" scenarios through what...
https://openreview.net/pdf/8e8a8d5dafd24c9cba49d3671b2ee34d0decdecf.pdf
Heteroscedastic Temporal Variational Autoencoder For Irregularly Sampled Time Series
https://openreview.net/forum?id=Az7opqbQE-3
https://openreview.net/forum?id=Az7opqbQE-3
Satya Narayan Shukla,Benjamin Marlin
ICLR 2022,Poster
Irregularly sampled time series commonly occur in several domains where they present a significant challenge to standard deep learning models. In this paper, we propose a new deep learning framework for probabilistic interpolation of irregularly sampled time series that we call the Heteroscedastic Temporal Variational ...
https://openreview.net/pdf/4a602866528e0ae9511889c65b61991ad9ddfd8b.pdf
A Neural Tangent Kernel Perspective of Infinite Tree Ensembles
https://openreview.net/forum?id=vUH85MOXO7h
https://openreview.net/forum?id=vUH85MOXO7h
Ryuichi Kanoh,Mahito Sugiyama
ICLR 2022,Poster
In practical situations, the tree ensemble is one of the most popular models along with neural networks. A soft tree is a variant of a decision tree. Instead of using a greedy method for searching splitting rules, the soft tree is trained using a gradient method in which the entire splitting operation is formulated in ...
https://openreview.net/pdf/39b3d2b8700abc51932e7eea69ff8d0868dc2be8.pdf
AlphaZero-based Proof Cost Network to Aid Game Solving
https://openreview.net/forum?id=nKWjE4QF1hB
https://openreview.net/forum?id=nKWjE4QF1hB
Ti-Rong Wu,Chung-Chin Shih,Ting Han Wei,Meng-Yu Tsai,Wei-Yuan Hsu,I-Chen Wu
ICLR 2022,Poster
The AlphaZero algorithm learns and plays games without hand-crafted expert knowledge. However, since its objective is to play well, we hypothesize that a better objective can be defined for the related but separate task of solving games. This paper proposes a novel approach to solving problems by modifying the training...
https://openreview.net/pdf/b5c23474ea991857d67e3e750bb82c36a669b2e9.pdf
Bayesian Framework for Gradient Leakage
https://openreview.net/forum?id=f2lrIbGx3x7
https://openreview.net/forum?id=f2lrIbGx3x7
Mislav Balunovic,Dimitar Iliev Dimitrov,Robin Staab,Martin Vechev
ICLR 2022,Poster
Federated learning is an established method for training machine learning models without sharing training data. However, recent work has shown that it cannot guarantee data privacy as shared gradients can still leak sensitive information. To formalize the problem of gradient leakage, we propose a theoretical framework ...
https://openreview.net/pdf/4e51a98c83f488bc5362a078c71216dab544be00.pdf
Universalizing Weak Supervision
https://openreview.net/forum?id=YpPiNigTzMT
https://openreview.net/forum?id=YpPiNigTzMT
Changho Shin,Winfred Li,Harit Vishwakarma,Nicholas Carl Roberts,Frederic Sala
ICLR 2022,Poster
Weak supervision (WS) frameworks are a popular way to bypass hand-labeling large datasets for training data-hungry models. These approaches synthesize multiple noisy but cheaply-acquired estimates of labels into a set of high-quality pseudo-labels for downstream training. However, the synthesis technique is specific to...
https://openreview.net/pdf/a2adc08eeb52dcddf2563c7bb42940946813b522.pdf
Maximum n-times Coverage for Vaccine Design
https://openreview.net/forum?id=ULfq0qR25dY
https://openreview.net/forum?id=ULfq0qR25dY
Ge Liu,Alexander Dimitrakakis,Brandon Carter,David Gifford
ICLR 2022,Poster
We introduce the maximum $n$-times coverage problem that selects $k$ overlays to maximize the summed coverage of weighted elements, where each element must be covered at least $n$ times. We also define the min-cost $n$-times coverage problem where the objective is to select the minimum set of overlays such that the sum...
https://openreview.net/pdf/9d61f13ecd3d02a7e3ed6243e5e82f05c5f456cf.pdf
KL Guided Domain Adaptation
https://openreview.net/forum?id=0JzqUlIVVDd
https://openreview.net/forum?id=0JzqUlIVVDd
A. Tuan Nguyen,Toan Tran,Yarin Gal,Philip Torr,Atilim Gunes Baydin
ICLR 2022,Poster
Domain adaptation is an important problem and often needed for real-world applications. In this problem, instead of i.i.d. training and testing datapoints, we assume that the source (training) data and the target (testing) data have different distributions. With that setting, the empirical risk minimization training pr...
https://openreview.net/pdf/943a05167d50e4a4de4e6c043f7c7e6374502f72.pdf
From Stars to Subgraphs: Uplifting Any GNN with Local Structure Awareness
https://openreview.net/forum?id=Mspk_WYKoEH
https://openreview.net/forum?id=Mspk_WYKoEH
Lingxiao Zhao,Wei Jin,Leman Akoglu,Neil Shah
ICLR 2022,Poster
Message Passing Neural Networks (MPNNs) are a common type of Graph Neural Network (GNN), in which each node’s representation is computed recursively by aggregating representations (“messages”) from its immediate neighbors akin to a star-shaped pattern. MPNNs are appealing for being efficient and scalable, however their...
https://openreview.net/pdf/cc341ac588b917bee10fc4d5bb31b4a119b6108b.pdf
NETWORK INSENSITIVITY TO PARAMETER NOISE VIA PARAMETER ATTACK DURING TRAINING
https://openreview.net/forum?id=-8sBpe7rDiV
https://openreview.net/forum?id=-8sBpe7rDiV
Julian Büchel,Fynn Firouz Faber,Dylan Richard Muir
ICLR 2022,Poster
Neuromorphic neural network processors, in the form of compute-in-memory crossbar arrays of memristors, or in the form of subthreshold analog and mixed-signal ASICs, promise enormous advantages in compute density and energy efficiency for NN-based ML tasks. However, these technologies are prone to computational non-ide...
https://openreview.net/pdf/b7b77ce8535702dba33084aa20eb08cae53193f4.pdf
Gradient Importance Learning for Incomplete Observations
https://openreview.net/forum?id=fXHl76nO2AZ
https://openreview.net/forum?id=fXHl76nO2AZ
Qitong Gao,Dong Wang,Joshua David Amason,Siyang Yuan,Chenyang Tao,Ricardo Henao,Majda Hadziahmetovic,Lawrence Carin,Miroslav Pajic
ICLR 2022,Poster
Though recent works have developed methods that can generate estimates (or imputations) of the missing entries in a dataset to facilitate downstream analysis, most depend on assumptions that may not align with real-world applications and could suffer from poor performance in subsequent tasks such as classification. Thi...
https://openreview.net/pdf/77f82d36ef5cbde5647d6e9f7fb7dd38ce4e2a91.pdf
Do Users Benefit From Interpretable Vision? A User Study, Baseline, And Dataset
https://openreview.net/forum?id=v6s3HVjPerv
https://openreview.net/forum?id=v6s3HVjPerv
Leon Sixt,Martin Schuessler,Oana-Iuliana Popescu,Philipp Weiß,Tim Landgraf
ICLR 2022,Poster
A variety of methods exist to explain image classification models. However, whether they provide any benefit to users over simply comparing various inputs and the model’s respective predictions remains unclear. We conducted a user study (N=240) to test how such a baseline explanation technique performs against concept-...
https://openreview.net/pdf/49e3023b785924a7159ee756c546ac2ec523e8ea.pdf
Understanding the Variance Collapse of SVGD in High Dimensions
https://openreview.net/forum?id=Qycd9j5Qp9J
https://openreview.net/forum?id=Qycd9j5Qp9J
Jimmy Ba,Murat A Erdogdu,Marzyeh Ghassemi,Shengyang Sun,Taiji Suzuki,Denny Wu,Tianzong Zhang
ICLR 2022,Poster
Stein variational gradient descent (SVGD) is a deterministic inference algorithm that evolves a set of particles to fit a target distribution. Despite its computational efficiency, SVGD often underestimates the variance of the target distribution in high dimensions. In this work we attempt to explain the variance colla...
https://openreview.net/pdf/71e77dab5447ab6226d0f2e58132575f2217dc3b.pdf
Generalisation in Lifelong Reinforcement Learning through Logical Composition
https://openreview.net/forum?id=ZOcX-eybqoL
https://openreview.net/forum?id=ZOcX-eybqoL
Geraud Nangue Tasse,Steven James,Benjamin Rosman
ICLR 2022,Poster
We leverage logical composition in reinforcement learning to create a framework that enables an agent to autonomously determine whether a new task can be immediately solved using its existing abilities, or whether a task-specific skill should be learned. In the latter case, the proposed algorithm also enables the agent...
https://openreview.net/pdf/89cb79a9b9bb6a9a833a7a8ae73c8c5a87792970.pdf
PolyLoss: A Polynomial Expansion Perspective of Classification Loss Functions
https://openreview.net/forum?id=gSdSJoenupI
https://openreview.net/forum?id=gSdSJoenupI
Zhaoqi Leng,Mingxing Tan,Chenxi Liu,Ekin Dogus Cubuk,Jay Shi,Shuyang Cheng,Dragomir Anguelov
ICLR 2022,Poster
Cross-entropy loss and focal loss are the most common choices when training deep neural networks for classification problems. Generally speaking, however, a good loss function can take on much more flexible forms, and should be tailored for different tasks and datasets. Motivated by how functions can be approximated vi...
https://openreview.net/pdf/d1430448cff98fb37273293f39735ba9c6a4313a.pdf
Improving Non-Autoregressive Translation Models Without Distillation
https://openreview.net/forum?id=I2Hw58KHp8O
https://openreview.net/forum?id=I2Hw58KHp8O
Xiao Shi Huang,Felipe Perez,Maksims Volkovs
ICLR 2022,Poster
Transformer-based autoregressive (AR) machine translation models have achieved significant performance improvements, nearing human-level accuracy on some languages. The AR framework translates one token at a time which can be time consuming, especially for long sequences. To accelerate inference, recent work has been e...
https://openreview.net/pdf/fe5e18c9939f10295c39693c81d77b03816cad63.pdf
A Theory of Tournament Representations
https://openreview.net/forum?id=zzk231Ms1Ih
https://openreview.net/forum?id=zzk231Ms1Ih
Arun Rajkumar,Vishnu Veerathu,Abdul Bakey Mir
ICLR 2022,Poster
Real-world tournaments are almost always intransitive. Recent works have noted that parametric models which assume $d$ dimensional node representations can effectively model intransitive tournaments. However, nothing is known about the structure of the class of tournaments that arise out of any fixed $d$ dimensional r...
https://openreview.net/pdf/a7853d8c301f8a37bc858f4c428d73862dabff26.pdf
Convergent and Efficient Deep Q Learning Algorithm
https://openreview.net/forum?id=OJm3HZuj4r7
https://openreview.net/forum?id=OJm3HZuj4r7
Zhikang T. Wang,Masahito Ueda
ICLR 2022,Poster
Despite the empirical success of the deep Q network (DQN) reinforcement learning algorithm and its variants, DQN is still not well understood and it does not guarantee convergence. In this work, we show that DQN can indeed diverge and cease to operate in realistic settings. Although there exist gradient-based convergen...
https://openreview.net/pdf/d999c3cb704da4722ea5330b5dd48600eb9c4ef4.pdf
Trigger Hunting with a Topological Prior for Trojan Detection
https://openreview.net/forum?id=TXsjU8BaibT
https://openreview.net/forum?id=TXsjU8BaibT
Xiaoling Hu,Xiao Lin,Michael Cogswell,Yi Yao,Susmit Jha,Chao Chen
ICLR 2022,Poster
Despite their success and popularity, deep neural networks (DNNs) are vulnerable when facing backdoor attacks. This impedes their wider adoption, especially in mission critical applications. This paper tackles the problem of Trojan detection, namely, identifying Trojaned models – models trained with poisoned data. One ...
https://openreview.net/pdf/4db1d42d467c296c5ec7fa3f38e37dcb5c140e84.pdf
Who Is the Strongest Enemy? Towards Optimal and Efficient Evasion Attacks in Deep RL
https://openreview.net/forum?id=JM2kFbJvvI
https://openreview.net/forum?id=JM2kFbJvvI
Yanchao Sun,Ruijie Zheng,Yongyuan Liang,Furong Huang
ICLR 2022,Poster
Evaluating the worst-case performance of a reinforcement learning (RL) agent under the strongest/optimal adversarial perturbations on state observations (within some constraints) is crucial for understanding the robustness of RL agents. However, finding the optimal adversary is challenging, in terms of both whether we ...
https://openreview.net/pdf/b11335ea1d1d4ca95531723261e11735e0550bc4.pdf
Chunked Autoregressive GAN for Conditional Waveform Synthesis
https://openreview.net/forum?id=v3aeIsY_vVX
https://openreview.net/forum?id=v3aeIsY_vVX
Max Morrison,Rithesh Kumar,Kundan Kumar,Prem Seetharaman,Aaron Courville,Yoshua Bengio
ICLR 2022,Poster
Conditional waveform synthesis models learn a distribution of audio waveforms given conditioning such as text, mel-spectrograms, or MIDI. These systems employ deep generative models that model the waveform via either sequential (autoregressive) or parallel (non-autoregressive) sampling. Generative adversarial networks ...
https://openreview.net/pdf/070239829c83980ec499e2eff346d48eafe3ecb5.pdf
COPA: Certifying Robust Policies for Offline Reinforcement Learning against Poisoning Attacks
https://openreview.net/forum?id=psh0oeMSBiF
https://openreview.net/forum?id=psh0oeMSBiF
Fan Wu,Linyi Li,Huan Zhang,Bhavya Kailkhura,Krishnaram Kenthapadi,Ding Zhao,Bo Li
ICLR 2022,Poster
As reinforcement learning (RL) has achieved near human-level performance in a variety of tasks, its robustness has raised great attention. While a vast body of research has explored test-time (evasion) attacks in RL and corresponding defenses, its robustness against training-time (poisoning) attacks remains largely una...
https://openreview.net/pdf/0a24a116cb24a1e99cd715566dae243e36472472.pdf
ExT5: Towards Extreme Multi-Task Scaling for Transfer Learning
https://openreview.net/forum?id=Vzh1BFUCiIX
https://openreview.net/forum?id=Vzh1BFUCiIX
Vamsi Aribandi,Yi Tay,Tal Schuster,Jinfeng Rao,Huaixiu Steven Zheng,Sanket Vaibhav Mehta,Honglei Zhuang,Vinh Q. Tran,Dara Bahri,Jianmo Ni,Jai Gupta,Kai Hui,Sebastian Ruder,Donald Metzler
ICLR 2022,Poster
Despite the recent success of multi-task learning and transfer learning for natural language processing (NLP), few works have systematically studied the effect of scaling up the number of tasks during pre-training. Towards this goal, this paper introduces ExMix (Extreme Mixture): a massive collection of 107 supervised ...
https://openreview.net/pdf/b64da5c159b90bf56d174fc67459b74928711232.pdf
Provable Adaptation across Multiway Domains via Representation Learning
https://openreview.net/forum?id=gRCCdgpVZf
https://openreview.net/forum?id=gRCCdgpVZf
Zhili Feng,Shaobo Han,Simon Shaolei Du
ICLR 2022,Poster
This paper studies zero-shot domain adaptation where each domain is indexed on a multi-dimensional array, and we only have data from a small subset of domains. Our goal is to produce predictors that perform well on \emph{unseen} domains. We propose a model which consists of a domain-invariant latent representation laye...
https://openreview.net/pdf/097cce8a39240bc2a614483e1cb4e0314237f10a.pdf
Efficient Token Mixing for Transformers via Adaptive Fourier Neural Operators
https://openreview.net/forum?id=EXHG-A3jlM
https://openreview.net/forum?id=EXHG-A3jlM
John Guibas,Morteza Mardani,Zongyi Li,Andrew Tao,Anima Anandkumar,Bryan Catanzaro
ICLR 2022,Poster
Vision transformers have delivered tremendous success in representation learning. This is primarily due to effective token mixing through self attention. However, this scales quadratically with the number of pixels, which becomes infeasible for high-resolution inputs. To cope with this challenge, we propose Adaptive Fo...
https://openreview.net/pdf/bec7c123720932f2545dfb12e85bab8ac5cca6ff.pdf
Sample Selection with Uncertainty of Losses for Learning with Noisy Labels
https://openreview.net/forum?id=xENf4QUL4LW
https://openreview.net/forum?id=xENf4QUL4LW
Xiaobo Xia,Tongliang Liu,Bo Han,Mingming Gong,Jun Yu,Gang Niu,Masashi Sugiyama
ICLR 2022,Poster
In learning with noisy labels, the sample selection approach is very popular, which regards small-loss data as correctly labeled data during training. However, losses are generated on-the-fly based on the model being trained with noisy labels, and thus large-loss data are likely but not certain to be incorrect. There ar...
https://openreview.net/pdf/0ebab5bba4b36eec025abfd2e21f947e05d6e662.pdf
Data-Driven Offline Optimization for Architecting Hardware Accelerators
https://openreview.net/forum?id=GsH-K1VIyy
https://openreview.net/forum?id=GsH-K1VIyy
Aviral Kumar,Amir Yazdanbakhsh,Milad Hashemi,Kevin Swersky,Sergey Levine
ICLR 2022,Poster
To attain higher efficiency, the industry has gradually reformed towards application-specific hardware accelerators. While such a paradigm shift is already starting to show promising results, designers need to spend considerable manual effort and perform large number of time-consuming simulations to find accelerators t...
https://openreview.net/pdf/62fa3ad6648729230b552447a872cf6777743905.pdf
Multi-Agent MDP Homomorphic Networks
https://openreview.net/forum?id=H7HDG--DJF0
https://openreview.net/forum?id=H7HDG--DJF0
Elise van der Pol,Herke van Hoof,Frans A Oliehoek,Max Welling
ICLR 2022,Poster
This paper introduces Multi-Agent MDP Homomorphic Networks, a class of networks that allows distributed execution using only local information, yet is able to share experience between global symmetries in the joint state-action space of cooperative multi-agent systems. In cooperative multi-agent systems, complex symmet...
https://openreview.net/pdf/3a8f28592a8f20859b54c37f57cb659f7b0664fa.pdf
Geometry-Consistent Neural Shape Representation with Implicit Displacement Fields
https://openreview.net/forum?id=yhCp5RcZD7
https://openreview.net/forum?id=yhCp5RcZD7
Wang Yifan,Lukas Rahmann,Olga Sorkine-hornung
ICLR 2022,Poster
We present implicit displacement fields, a novel representation for detailed 3D geometry. Inspired by a classic surface deformation technique, displacement mapping, our method represents a complex surface as a smooth base surface plus a displacement along the base's normal directions, resulting in a frequency-based sha...
https://openreview.net/pdf/55c1560b8382311a7f02b90aaba2fa21e4475e9d.pdf
Modeling Label Space Interactions in Multi-label Classification using Box Embeddings
https://openreview.net/forum?id=tyTH9kOxcvh
https://openreview.net/forum?id=tyTH9kOxcvh
Dhruvesh Patel,Pavitra Dangati,Jay-Yoon Lee,Michael Boratko,Andrew McCallum
ICLR 2022,Poster
Multi-label classification is a challenging structured prediction task in which a set of output class labels are predicted for each input. Real-world datasets often have natural or latent taxonomic relationships between labels, making it desirable for models to employ label representations capable of capturing such tax...
https://openreview.net/pdf/f5671d43125692a6533d9c7a1996335b8a1cd482.pdf
It Takes Two to Tango: Mixup for Deep Metric Learning
https://openreview.net/forum?id=ZKy2X3dgPA
https://openreview.net/forum?id=ZKy2X3dgPA
Shashanka Venkataramanan,Bill Psomas,Ewa Kijak,laurent amsaleg,Konstantinos Karantzalos,Yannis Avrithis
ICLR 2022,Poster
Metric learning involves learning a discriminative representation such that embeddings of similar classes are encouraged to be close, while embeddings of dissimilar classes are pushed far apart. State-of-the-art methods focus mostly on sophisticated loss functions or mining strategies. On the one hand, metric learning ...
https://openreview.net/pdf/1b4683c706bc39fb7b56b3982f8c10166b29773d.pdf
Data Efficient Language-Supervised Zero-Shot Recognition with Optimal Transport Distillation
https://openreview.net/forum?id=G89-1yZLFHk
https://openreview.net/forum?id=G89-1yZLFHk
Bichen Wu,Ruizhe Cheng,Peizhao Zhang,Tianren Gao,Joseph E. Gonzalez,Peter Vajda
ICLR 2022,Poster
Traditional computer vision models are trained to predict a fixed set of predefined categories. Recently, natural language has been shown to be a broader and richer source of supervision that provides finer descriptions to visual concepts than supervised "gold" labels. Previous works, such as CLIP, use InfoNCE loss to ...
https://openreview.net/pdf/4692c27fcf85afed7f22e02ea4a1c14104fce2a4.pdf
A Statistical Framework for Efficient Out of Distribution Detection in Deep Neural Networks
https://openreview.net/forum?id=Oy9WeuZD51
https://openreview.net/forum?id=Oy9WeuZD51
Matan Haroush,Tzviel Frostig,Ruth Heller,Daniel Soudry
ICLR 2022,Poster
Background. Commonly, Deep Neural Networks (DNNs) generalize well on samples drawn from a distribution similar to that of the training set. However, DNNs' predictions are brittle and unreliable when the test samples are drawn from a dissimilar distribution. This is a major concern for deployment in real-world applicati...
https://openreview.net/pdf/8ab4fc0f10bb1b17497961ee8ff9912af8ed2cc3.pdf
FedBABU: Toward Enhanced Representation for Federated Image Classification
https://openreview.net/forum?id=HuaYQfggn5u
https://openreview.net/forum?id=HuaYQfggn5u
Jaehoon Oh,SangMook Kim,Se-Young Yun
ICLR 2022,Poster
Federated learning has evolved to improve a single global model under data heterogeneity (as a curse) or to develop multiple personalized models using data heterogeneity (as a blessing). However, little research has considered both directions simultaneously. In this paper, we first investigate the relationship between ...
https://openreview.net/pdf/09e0b377fa4e3200e80d267b3e1df94235e10a45.pdf
Should I Run Offline Reinforcement Learning or Behavioral Cloning?
https://openreview.net/forum?id=AP1MKT37rJ
https://openreview.net/forum?id=AP1MKT37rJ
Aviral Kumar,Joey Hong,Anikait Singh,Sergey Levine
ICLR 2022,Poster
Offline reinforcement learning (RL) algorithms can acquire effective policies by utilizing only previously collected experience, without any online interaction. While it is widely understood that offline RL is able to extract good policies even from highly suboptimal data, in practice offline RL is often used with dat...
https://openreview.net/pdf/ab91050974b19858a9a241236b4d69019903de0e.pdf
Learning State Representations via Retracing in Reinforcement Learning
https://openreview.net/forum?id=CLpxpXqqBV
https://openreview.net/forum?id=CLpxpXqqBV
Changmin Yu,Dong Li,Jianye HAO,Jun Wang,Neil Burgess
ICLR 2022,Poster
We propose learning via retracing, a novel self-supervised approach for learning the state representation (and the associated dynamics model) for reinforcement learning tasks. In addition to the predictive (reconstruction) supervision in the forward direction, we propose to include "retraced" transitions for representa...
https://openreview.net/pdf/04d24e2870546f3dcff312162e1b4006ecd641b7.pdf
Open-World Semi-Supervised Learning
https://openreview.net/forum?id=O-r8LOR-CCA
https://openreview.net/forum?id=O-r8LOR-CCA
Kaidi Cao,Maria Brbic,Jure Leskovec
ICLR 2022,Poster
A fundamental limitation of applying semi-supervised learning in real-world settings is the assumption that unlabeled test data contains only classes previously encountered in the labeled training data. However, this assumption rarely holds for data in-the-wild, where instances belonging to novel classes may appear at ...
https://openreview.net/pdf/e5ffbb438b307d601bd7794c87fae3c23950a63f.pdf
Evading Adversarial Example Detection Defenses with Orthogonal Projected Gradient Descent
https://openreview.net/forum?id=af1eUDdUVz
https://openreview.net/forum?id=af1eUDdUVz
Oliver Bryniarski,Nabeel Hingun,Pedro Pachuca,Vincent Wang,Nicholas Carlini
ICLR 2022,Poster
Evading adversarial example detection defenses requires finding adversarial examples that must simultaneously (a) be misclassified by the model and (b) be detected as non-adversarial. We find that existing attacks that attempt to satisfy multiple simultaneous constraints often over-optimize against one constraint at th...
https://openreview.net/pdf/3d2eb96b012475581aa80cda16373c217e28c087.pdf
Reducing Excessive Margin to Achieve a Better Accuracy vs. Robustness Trade-off
https://openreview.net/forum?id=Azh9QBQ4tR7
https://openreview.net/forum?id=Azh9QBQ4tR7
Rahul Rade,Seyed-Mohsen Moosavi-Dezfooli
ICLR 2022,Poster
While adversarial training has become the de facto approach for training robust classifiers, it leads to a drop in accuracy. This has led to prior works postulating that accuracy is inherently at odds with robustness. Yet, the phenomenon remains inexplicable. In this paper, we closely examine the changes induced in the...
https://openreview.net/pdf/c2a72787c4e6f0d24586b17eab7ca97027346386.pdf
Expressivity of Emergent Languages is a Trade-off between Contextual Complexity and Unpredictability
https://openreview.net/forum?id=WxuE_JWxjkW
https://openreview.net/forum?id=WxuE_JWxjkW
Shangmin Guo,Yi Ren,Kory Wallace Mathewson,Simon Kirby,Stefano V Albrecht,Kenny Smith
ICLR 2022,Poster
Researchers are using deep learning models to explore the emergence of language in various language games, where agents interact and develop an emergent language to solve tasks. We focus on the factors that determine the expressivity of emergent languages, which reflects the amount of information about input spaces tho...
https://openreview.net/pdf/be46689741877d2b59dc56c09443500af7dd2941.pdf
Fast AdvProp
https://openreview.net/forum?id=hcoswsDHNAW
https://openreview.net/forum?id=hcoswsDHNAW
Jieru Mei,Yucheng Han,Yutong Bai,Yixiao Zhang,Yingwei Li,Xianhang Li,Alan Yuille,Cihang Xie
ICLR 2022,Poster
Adversarial Propagation (AdvProp) is an effective way to improve recognition models, leveraging adversarial examples. Nonetheless, AdvProp suffers from the extremely slow training speed, mainly because: a) extra forward and backward passes are required for generating adversarial examples; b) both original samples and t...
https://openreview.net/pdf/12e365a996eeb801b2173df149f6f8bc69ec02fa.pdf
Triangle and Four Cycle Counting with Predictions in Graph Streams
https://openreview.net/forum?id=8in_5gN9I0
https://openreview.net/forum?id=8in_5gN9I0
Justin Y Chen,Talya Eden,Piotr Indyk,Honghao Lin,Shyam Narayanan,Ronitt Rubinfeld,Sandeep Silwal,Tal Wagner,David Woodruff,Michael Zhang
ICLR 2022,Poster
We propose data-driven one-pass streaming algorithms for estimating the number of triangles and four cycles, two fundamental problems in graph analytics that are widely studied in the graph data stream literature. Recently, Hsu et al. (2019) and Jiang et al. (2020) applied machine learning techniques in other data stre...
https://openreview.net/pdf/25b70c42018200ce5f79c1f1dfc16f4c95ff9304.pdf
Is Fairness Only Metric Deep? Evaluating and Addressing Subgroup Gaps in Deep Metric Learning
https://openreview.net/forum?id=js62_xuLDDv
https://openreview.net/forum?id=js62_xuLDDv
Natalie Dullerud,Karsten Roth,Kimia Hamidieh,Nicolas Papernot,Marzyeh Ghassemi
ICLR 2022,Poster
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 fir...
https://openreview.net/pdf/f404cf882e197b2c86f3e62a769c3cbf9024a9b5.pdf
NodePiece: Compositional and Parameter-Efficient Representations of Large Knowledge Graphs
https://openreview.net/forum?id=xMJWUKJnFSw
https://openreview.net/forum?id=xMJWUKJnFSw
Mikhail Galkin,Etienne Denis,Jiapeng Wu,William L. Hamilton
ICLR 2022,Poster
Conventional representation learning algorithms for knowledge graphs (KG) map each entity to a unique embedding vector. Such a shallow lookup results in a linear growth of memory consumption for storing the embedding matrix and incurs high computational costs of working with real-world KGs. Drawing parallels with subw...
https://openreview.net/pdf/6eb641d163812ce838dbad1b8e7fddebb2c72c12.pdf
Pix2seq: A Language Modeling Framework for Object Detection
https://openreview.net/forum?id=e42KbIw6Wb
https://openreview.net/forum?id=e42KbIw6Wb
Ting Chen,Saurabh Saxena,Lala Li,David J. Fleet,Geoffrey Hinton
ICLR 2022,Poster
We present Pix2Seq, a simple and generic framework for object detection. Unlike existing approaches that explicitly integrate prior knowledge about the task, we cast object detection as a language modeling task conditioned on the observed pixel inputs. Object descriptions (e.g., bounding boxes and class labels) are exp...
https://openreview.net/pdf/1f7291d96e3b195bdf0664dfb0f5313b0eab7a04.pdf
Particle Stochastic Dual Coordinate Ascent: Exponential convergent algorithm for mean field neural network optimization
https://openreview.net/forum?id=PQQp7AJwz3
https://openreview.net/forum?id=PQQp7AJwz3
Kazusato Oko,Taiji Suzuki,Atsushi Nitanda,Denny Wu
ICLR 2022,Poster
We introduce Particle-SDCA, a gradient-based optimization algorithm for two-layer neural networks in the mean field regime that achieves exponential convergence rate in regularized empirical risk minimization. The proposed algorithm can be regarded as an infinite dimensional extension of Stochastic Dual Coordinate Asce...
https://openreview.net/pdf/b6a0af59072ab41c5553c6952e5a786b25d0adde.pdf
The Effects of Invertibility on the Representational Complexity of Encoders in Variational Autoencoders
https://openreview.net/forum?id=7_JR7WpwKV1
https://openreview.net/forum?id=7_JR7WpwKV1
Divyansh Pareek,Andrej Risteski
ICLR 2022,Poster
Training and using modern neural-network based latent-variable generative models (like Variational Autoencoders) often require simultaneously training a generative direction along with an inferential (encoding) direction, which approximates the posterior distribution over the latent variables. Thus, the question arises...
https://openreview.net/pdf/4116475bedc76111284bad627cb9a8fbaec2059b.pdf
Tracking the risk of a deployed model and detecting harmful distribution shifts
https://openreview.net/forum?id=Ro_zAjZppv
https://openreview.net/forum?id=Ro_zAjZppv
Aleksandr Podkopaev,Aaditya Ramdas
ICLR 2022,Poster
When deployed in the real world, machine learning models inevitably encounter changes in the data distribution, and certain---but not all---distribution shifts could result in significant performance degradation. In practice, it may make sense to ignore benign shifts, under which the performance of a deployed model doe...
https://openreview.net/pdf/f763a5271b61d98bca4127ab14ce483150d152c4.pdf
Towards Understanding the Robustness Against Evasion Attack on Categorical Data
https://openreview.net/forum?id=BmJV7kyAmg
https://openreview.net/forum?id=BmJV7kyAmg
Hongyan Bao,Yufei Han,Yujun Zhou,Yun Shen,Xiangliang Zhang
ICLR 2022,Poster
Characterizing and assessing the adversarial vulnerability of classification models with categorical input has been a practically important, while rarely explored research problem. Our work echoes the challenge by first unveiling the impact factors of adversarial vulnerability of classification models with categorical ...
https://openreview.net/pdf/b599972b615dea56e3cd777bb3c09e18b73ba736.pdf
Learning Curves for SGD on Structured Features
https://openreview.net/forum?id=WPI2vbkAl3Q
https://openreview.net/forum?id=WPI2vbkAl3Q
Blake Bordelon,Cengiz Pehlevan
ICLR 2022,Poster
The generalization performance of a machine learning algorithm such as a neural network depends in a non-trivial way on the structure of the data distribution. To analyze the influence of data structure on test loss dynamics, we study an exactly solveable model of stochastic gradient descent (SGD) on the square loss wh...
https://openreview.net/pdf/05e1bd43845bd2321a0ab8593b8960931a65e24e.pdf
NASViT: Neural Architecture Search for Efficient Vision Transformers with Gradient Conflict aware Supernet Training
https://openreview.net/forum?id=Qaw16njk6L
https://openreview.net/forum?id=Qaw16njk6L
Chengyue Gong,Dilin Wang,Meng Li,Xinlei Chen,Zhicheng Yan,Yuandong Tian,qiang liu,Vikas Chandra
ICLR 2022,Poster
Designing accurate and efficient vision transformers (ViTs) is a highly important but challenging task. Supernet-based one-shot neural architecture search (NAS) enables fast architecture optimization and has achieved state-of-the-art (SOTA) results on convolutional neural networks (CNNs). However, directly applying the...
https://openreview.net/pdf/a6df48abb7e0bb493e7c343c46beb7b365cdc788.pdf
Graphon based Clustering and Testing of Networks: Algorithms and Theory
https://openreview.net/forum?id=sTNHCrIKDQc
https://openreview.net/forum?id=sTNHCrIKDQc
Mahalakshmi Sabanayagam,Leena Chennuru Vankadara,Debarghya Ghoshdastidar
ICLR 2022,Poster
Network-valued data are encountered in a wide range of applications, and pose challenges in learning due to their complex structure and absence of vertex correspondence. Typical examples of such problems include classification or grouping of protein structures and social networks. Various methods, ranging from graph ke...
https://openreview.net/pdf/bc3a82e090f7f3cfaa9a92ef69181887e0348ede.pdf
Network Augmentation for Tiny Deep Learning
https://openreview.net/forum?id=TYw3-OlrRm-
https://openreview.net/forum?id=TYw3-OlrRm-
Han Cai,Chuang Gan,Ji Lin,Song Han
ICLR 2022,Poster
We introduce Network Augmentation (NetAug), a new training method for improving the performance of tiny neural networks. Existing regularization techniques (e.g., data augmentation, dropout) have shown much success on large neural networks by adding noise to overcome over-fitting. However, we found these techniques hur...
https://openreview.net/pdf/484496875b902e745fc4d6514abb817e7be477c2.pdf
Bridging the Gap: Providing Post-Hoc Symbolic Explanations for Sequential Decision-Making Problems with Inscrutable Representations
https://openreview.net/forum?id=o-1v9hdSult
https://openreview.net/forum?id=o-1v9hdSult
Sarath Sreedharan,Utkarsh Soni,Mudit Verma,Siddharth Srivastava,Subbarao Kambhampati
ICLR 2022,Poster
As increasingly complex AI systems are introduced into our daily lives, it becomes important for such systems to be capable of explaining the rationale for their decisions and allowing users to contest these decisions. A significant hurdle to allowing for such explanatory dialogue could be the {\em vocabulary mismatch}...
https://openreview.net/pdf/2558c3735ba361f65aac84ecf8e9f4624e87dec8.pdf
Distributional Reinforcement Learning with Monotonic Splines
https://openreview.net/forum?id=C8Ltz08PtBp
https://openreview.net/forum?id=C8Ltz08PtBp
Yudong Luo,Guiliang Liu,Haonan Duan,Oliver Schulte,Pascal Poupart
ICLR 2022,Poster
Distributional Reinforcement Learning (RL) differs from traditional RL by estimating the distribution over returns to capture the intrinsic uncertainty of MDPs. One key challenge in distributional RL lies in how to parameterize the quantile function when minimizing the Wasserstein metric of temporal differences. Existi...
https://openreview.net/pdf/376a906de470631ee01098610befe6addc3d72de.pdf
Toward Faithful Case-based Reasoning through Learning Prototypes in a Nearest Neighbor-friendly Space.
https://openreview.net/forum?id=R79ZGjHhv6p
https://openreview.net/forum?id=R79ZGjHhv6p
Seyed Omid Davoudi,Majid Komeili
ICLR 2022,Poster
Recent advances in machine learning have brought opportunities for the ever-increasing use of AI in the real world. This has created concerns about the black-box nature of many of the most recent machine learning approaches. In this work, we propose an interpretable neural network that leverages metric and prototype le...
https://openreview.net/pdf/6d0714a184aa752df631ed2df558e8cfee0d4bb9.pdf
Augmented Sliced Wasserstein Distances
https://openreview.net/forum?id=iMqTLyfwnOO
https://openreview.net/forum?id=iMqTLyfwnOO
Xiongjie Chen,Yongxin Yang,Yunpeng Li
ICLR 2022,Poster
While theoretically appealing, the application of the Wasserstein distance to large-scale machine learning problems has been hampered by its prohibitive computational cost. The sliced Wasserstein distance and its variants improve the computational efficiency through the random projection, yet they suffer from low accur...
https://openreview.net/pdf/d09a765a0ca6e8fe66e61db6af5518d089814c41.pdf
Relational Learning with Variational Bayes
https://openreview.net/forum?id=Az-7gJc6lpr
https://openreview.net/forum?id=Az-7gJc6lpr
Kuang-Hung Liu
ICLR 2022,Poster
In psychology, relational learning refers to the ability to recognize and respond to relationship among objects irrespective of the nature of those objects. Relational learning has long been recognized as a hallmark of human cognition and a key question in artificial intelligence research. In this work, we propose an u...
https://openreview.net/pdf/9d3dfe42360aa203adb14bacece6acbb08064ac0.pdf
Provably Robust Adversarial Examples
https://openreview.net/forum?id=UMfhoMtIaP5
https://openreview.net/forum?id=UMfhoMtIaP5
Dimitar Iliev Dimitrov,Gagandeep Singh,Timon Gehr,Martin Vechev
ICLR 2022,Poster
We introduce the concept of provably robust adversarial examples for deep neural networks – connected input regions constructed from standard adversarial examples which are guaranteed to be robust to a set of real-world perturbations (such as changes in pixel intensity and geometric transformations). We present a novel...
https://openreview.net/pdf/3b8eb27fbc166f48033673d3fadc49a86ef0b79f.pdf
Joint Shapley values: a measure of joint feature importance
https://openreview.net/forum?id=vcUmUvQCloe
https://openreview.net/forum?id=vcUmUvQCloe
Chris Harris,Richard Pymar,Colin Rowat
ICLR 2022,Poster
The Shapley value is one of the most widely used measures of feature importance partly as it measures a feature's average effect on a model's prediction. We introduce joint Shapley values, which directly extend Shapley's axioms and intuitions: joint Shapley values measure a set of features' average effect on a model's...
https://openreview.net/pdf/7d8a95bb048b3b204b4a1c9a95e93486a12439a1.pdf
Low-Budget Active Learning via Wasserstein Distance: An Integer Programming Approach
https://openreview.net/forum?id=v8OlxjGn23S
https://openreview.net/forum?id=v8OlxjGn23S
Rafid Mahmood,Sanja Fidler,Marc T Law
ICLR 2022,Poster
Active learning is the process of training a model with limited labeled data by selecting a core subset of an unlabeled data pool to label. The large scale of data sets used in deep learning forces most sample selection strategies to employ efficient heuristics. This paper introduces an integer optimization problem for...
https://openreview.net/pdf/9dac127c30d4567d8dde179f21749b9ca5494686.pdf
Efficient Self-supervised Vision Transformers for Representation Learning
https://openreview.net/forum?id=fVu3o-YUGQK
https://openreview.net/forum?id=fVu3o-YUGQK
Chunyuan Li,Jianwei Yang,Pengchuan Zhang,Mei Gao,Bin Xiao,Xiyang Dai,Lu Yuan,Jianfeng Gao
ICLR 2022,Poster
This paper investigates two techniques for developing efficient self-supervised vision transformers (EsViT) for visual representation learning. First, we show through a comprehensive empirical study that multi-stage architectures with sparse self-attentions can significantly reduce modeling complexity but with a cost o...
https://openreview.net/pdf/e7b63dccef8ad598db1c36a2386c8d8a63058e8e.pdf
Visual Representation Learning Does Not Generalize Strongly Within the Same Domain
https://openreview.net/forum?id=9RUHPlladgh
https://openreview.net/forum?id=9RUHPlladgh
Lukas Schott,Julius Von Kügelgen,Frederik Träuble,Peter Vincent Gehler,Chris Russell,Matthias Bethge,Bernhard Schölkopf,Francesco Locatello,Wieland Brendel
ICLR 2022,Poster
An important component for generalization in machine learning is to uncover underlying latent factors of variation as well as the mechanism through which each factor acts in the world. In this paper, we test whether 17 unsupervised, weakly supervised, and fully supervised representation learning approaches correctly in...
https://openreview.net/pdf/775e024ab2e9ce40e6b2f7608d5b1eb2c1136e75.pdf
Hidden Convexity of Wasserstein GANs: Interpretable Generative Models with Closed-Form Solutions
https://openreview.net/forum?id=e2Lle5cij9D
https://openreview.net/forum?id=e2Lle5cij9D
Arda Sahiner,Tolga Ergen,Batu Ozturkler,Burak Bartan,John M. Pauly,Morteza Mardani,Mert Pilanci
ICLR 2022,Poster
Generative Adversarial Networks (GANs) are commonly used for modeling complex distributions of data. Both the generators and discriminators of GANs are often modeled by neural networks, posing a non-transparent optimization problem which is non-convex and non-concave over the generator and discriminator, respectively. ...
https://openreview.net/pdf/733796fc142ddb063afc1a0818ecba208aef1465.pdf
Memory Augmented Optimizers for Deep Learning
https://openreview.net/forum?id=NRX9QZ6yqt
https://openreview.net/forum?id=NRX9QZ6yqt
Paul-Aymeric Martin McRae,Prasanna Parthasarathi,Mido Assran,Sarath Chandar
ICLR 2022,Poster
Popular approaches for minimizing loss in data-driven learning often involve an abstraction or an explicit retention of the history of gradients for efficient parameter updates. The aggregated history of gradients nudges the parameter updates in the right direction even when the gradients at any given step are not inf...
https://openreview.net/pdf/874e2c95385be68f564d4d96107e652253f10706.pdf
Orchestrated Value Mapping for Reinforcement Learning
https://openreview.net/forum?id=c87d0TS4yX
https://openreview.net/forum?id=c87d0TS4yX
Mehdi Fatemi,Arash Tavakoli
ICLR 2022,Poster
We present a general convergent class of reinforcement learning algorithms that is founded on two distinct principles: (1) mapping value estimates to a different space using arbitrary functions from a broad class, and (2) linearly decomposing the reward signal into multiple channels. The first principle enables incorpo...
https://openreview.net/pdf/9ef3cef089b9f45f5bdb93fddb0ed8ccfa9e3268.pdf
Learning to Generalize across Domains on Single Test Samples
https://openreview.net/forum?id=CIaQKbTBwtU
https://openreview.net/forum?id=CIaQKbTBwtU
Zehao Xiao,Xiantong Zhen,Ling Shao,Cees G. M. Snoek
ICLR 2022,Poster
We strive to learn a model from a set of source domains that generalizes well to unseen target domains. The main challenge in such a domain generalization scenario is the unavailability of any target domain data during training, resulting in the learned model not being explicitly adapted to the unseen target domains. W...
https://openreview.net/pdf/4fcc67594340f12c1beb7e4f1ce64c7be6f70c0a.pdf
Prototype memory and attention mechanisms for few shot image generation
https://openreview.net/forum?id=lY0-7bj0Vfz
https://openreview.net/forum?id=lY0-7bj0Vfz
Tianqin Li,Zijie Li,Andrew Luo,Harold Rockwell,Amir Barati Farimani,Tai Sing Lee
ICLR 2022,Poster
Recent discoveries indicate that the neural codes in the primary visual cortex (V1) of macaque monkeys are complex, diverse and sparse. This leads us to ponder the computational advantages and functional role of these “grandmother cells." Here, we propose that such cells can serve as prototype memory priors that bias a...
https://openreview.net/pdf/c2a4a72f1bd5890c4beeb93de11cac4746eae2c1.pdf
TPU-GAN: Learning temporal coherence from dynamic point cloud sequences
https://openreview.net/forum?id=FEBFJ98FKx
https://openreview.net/forum?id=FEBFJ98FKx
Zijie Li,Tianqin Li,Amir Barati Farimani
ICLR 2022,Poster
Point cloud sequence is an important data representation that provides flexible shape and motion information. Prior work demonstrates that incorporating scene flow information into loss can make model learn temporally coherent feature spaces. However, it is prohibitively expensive to acquire point correspondence inform...
https://openreview.net/pdf/52569840ae5698d2203efde4f8f06d012fa7868a.pdf
A First-Occupancy Representation for Reinforcement Learning
https://openreview.net/forum?id=JBAZe2yN6Ub
https://openreview.net/forum?id=JBAZe2yN6Ub
Ted Moskovitz,Spencer R Wilson,Maneesh Sahani
ICLR 2022,Poster
Both animals and artificial agents benefit from state representations that support rapid transfer of learning across tasks and which enable them to efficiently traverse their environments to reach rewarding states. The successor representation (SR), which measures the expected cumulative, discounted state occupancy un...
https://openreview.net/pdf/46abdff2d131f44012d855cdd93c0fa7034d601a.pdf
Deep ReLU Networks Preserve Expected Length
https://openreview.net/forum?id=ci7LBzDn2Q
https://openreview.net/forum?id=ci7LBzDn2Q
Boris Hanin,Ryan Jeong,David Rolnick
ICLR 2022,Poster
Assessing the complexity of functions computed by a neural network helps us understand how the network will learn and generalize. One natural measure of complexity is how the network distorts length - if the network takes a unit-length curve as input, what is the length of the resulting curve of outputs? It has been wi...
https://openreview.net/pdf/726f7b1d7efcb38a8f1685099dbfc32c938b1267.pdf