ICLR

Towards Understanding the Data Dependency of Mixup-style Training

Muthu Chidambaram, Xiang Wang, Yuzheng Hu, Chenwei Wu, Rong Ge

Comments 26 pages, 14 figures, Accepted to ICLR 2022 (Spotlight)

Subspace Regularizers for Few-Shot Class Incremental Learning

Afra Feyza Akyürek, Ekin Akyürek, Derry Tanti Wijaya, Jacob Andreas

Comments ICLR 2022. Code is available through https://github.com/feyzaakyurek/subspace-reg

Sqrt(d) Dimension Dependence of Langevin Monte Carlo

Ruilin Li, Hongyuan Zha, Molei Tao

Comments v1 submitted on May 28, 2021 (NeurIPS 2021 deadline); v2 added an important reference and discussions; v3 is the camera ready version

Journal ref ICLR 2022

DKM: Differentiable K-Means Clustering Layer for Neural Network Compression

Minsik Cho, Keivan A. Vahid, Saurabh Adya, Mohammad Rastegari

Comments ICLR 2022

PriorGrad: Improving Conditional Denoising Diffusion Models with Data-Dependent Adaptive Prior

Sang-gil Lee, Heeseung Kim, Chaehun Shin, Xu Tan, Chang Liu, Qi Meng, Tao Qin, Wei Chen, Sungroh Yoon, Tie-Yan Liu

Comments ICLR 2022. 19 pages, 7 figures, 8 tables. Audio samples: https://speechresearch.github.io/priorgrad/

Gaussian Mixture Convolution Networks

Adam Celarek, Pedro Hermosilla, Bernhard Kerbl, Timo Ropinski, Michael Wimmer

Comments To be published in ICLR 2022

Graph Auto-Encoder Via Neighborhood Wasserstein Reconstruction

Mingyue Tang, Carl Yang, Pan Li

Comments ICLR 2022; Code available at https://github.com/mtang724/NWR-GAE

Learning Weakly-Supervised Contrastive Representations

Yao-Hung Hubert Tsai, Tianqin Li, Weixin Liu, Peiyuan Liao, Ruslan Salakhutdinov, Louis-Philippe Morency

Comments Published as ICLR 2022. arXiv admin note: substantial text overlap with arXiv:2106.02869

Learning Graphon Mean Field Games and Approximate Nash Equilibria

Kai Cui, Heinz Koeppl

Comments Accepted to the Tenth International Conference on Learning Representations (ICLR); Fixed some typos

A global convergence theory for deep ReLU implicit networks via over-parameterization

Tianxiang Gao, Hailiang Liu, Jia Liu, Hridesh Rajan, Hongyang Gao

Comments Accepted by ICLR 2022

Multi-Level Local SGD for Heterogeneous Hierarchical Networks

Timothy Castiglia, Anirban Das, Stacy Patterson

Comments 36 pages, 10 figures, ICLR 2021

A Fair Comparison of Graph Neural Networks for Graph Classification

Federico Errica, Marco Podda, Davide Bacciu, Alessio Micheli

Comments Extended version of the paper published at the International Conference on Learning Representations (ICLR), 2020. Additional results are shown in the appendix

Revisiting Over-smoothing in BERT from the Perspective of Graph

Han Shi, Jiahui Gao, Hang Xu, Xiaodan Liang, Zhenguo Li, Lingpeng Kong, Stephen M. S. Lee, James T. Kwok

Comments Accepted by ICLR 2022 (Spotlight)

Signing the Supermask: Keep, Hide, Invert

Nils Koster, Oliver Grothe, Achim Rettinger

Comments ICLR 2022 camera ready

Towards Evaluating the Robustness of Neural Networks Learned by Transduction

Jiefeng Chen, Xi Wu, Yang Guo, Yingyu Liang, Somesh Jha

Comments Paper published at ICLR 2022. arXiv admin note: text overlap with arXiv:2106.08387

Learning Versatile Neural Architectures by Propagating Network Codes

Mingyu Ding, Yuqi Huo, Haoyu Lu, Linjie Yang, Zhe Wang, Zhiwu Lu, Jingdong Wang, Ping Luo

Comments ICLR 2022. Project page: https://network-propagation.github.io

Wisdom of Committees: An Overlooked Approach To Faster and More Accurate Models

Xiaofang Wang, Dan Kondratyuk, Eric Christiansen, Kris M. Kitani, Yair Alon, Elad Eban

Comments ICLR 2022

Objects in Semantic Topology

Shuo Yang, Peize Sun, Yi Jiang, Xiaobo Xia, Ruiheng Zhang, Zehuan Yuan, Changhu Wang, Ping Luo, Min Xu

Comments ICLR 2022

RotoGrad: Gradient Homogenization in Multitask Learning

Adrián Javaloy, Isabel Valera

Comments Spotlight at ICLR 2022. 24 pages, 9 figures

Vulnerability-Aware Poisoning Mechanism for Online RL with Unknown Dynamics

Yanchao Sun, Da Huo, Furong Huang

Journal ref The Ninth International Conference on Learning Representations (ICLR 2021)

SQuant: On-the-Fly Data-Free Quantization via Diagonal Hessian Approximation

Cong Guo, Yuxian Qiu, Jingwen Leng, Xiaotian Gao, Chen Zhang, Yunxin Liu, Fan Yang, Yuhao Zhu, Minyi Guo

Comments 18 pages, 2 figures, ICLR 2022

Imbedding Deep Neural Networks

Andrew Corbett, Dmitry Kangin

Comments Accepted as a spotlight paper at the 10th International Conference on Learning Representations (ICLR), 2022

Delaunay Component Analysis for Evaluation of Data Representations

Petra Poklukar, Vladislav Polianskii, Anastasia Varava, Florian Pokorny, Danica Kragic

Comments ICLR 2022 camera ready

Reinforcement Learning in Presence of Discrete Markovian Context Evolution

Hang Ren, Aivar Sootla, Taher Jafferjee, Junxiao Shen, Jun Wang, Haitham Bou-Ammar

Comments Accepted to ICLR 2022

MetaShift: A Dataset of Datasets for Evaluating Contextual Distribution Shifts and Training Conflicts

Weixin Liang, James Zou

Comments ICLR 2022. Code & data available at https://github.com/Weixin-Liang/MetaShift

Rethinking Goal-conditioned Supervised Learning and Its Connection to Offline RL

Rui Yang, Yiming Lu, Wenzhe Li, Hao Sun, Meng Fang, Yali Du, Xiu Li, Lei Han, Chongjie Zhang

Comments Accepted by International Conference on Learning Representations (ICLR) 2022

No Parameters Left Behind: Sensitivity Guided Adaptive Learning Rate for Training Large Transformer Models

Chen Liang, Haoming Jiang, Simiao Zuo, Pengcheng He, Xiaodong Liu, Jianfeng Gao, Weizhu Chen, Tuo Zhao

Comments Proceedings of ICLR 2022

How does unlabeled data improve generalization in self-training? A one-hidden-layer theoretical analysis

Shuai Zhang, Meng Wang, Sijia Liu, Pin-Yu Chen, Jinjun Xiong

Comments 36 pages

Journal ref Tenth International Conference on Learning Representations 2022

The Effects of Reward Misspecification: Mapping and Mitigating Misaligned Models

Alexander Pan, Kush Bhatia, Jacob Steinhardt

Comments ICLR 2022; 19 pages

Benchmarking the Spectrum of Agent Capabilities

Danijar Hafner

Comments Published at ICLR 2022. Website: https://danijar.com/crafter