{"id":725710,"date":"2021-02-11T15:37:17","date_gmt":"2021-02-11T23:37:17","guid":{"rendered":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/?post_type=msr-event&#038;p=725710"},"modified":"2025-08-06T11:51:54","modified_gmt":"2025-08-06T18:51:54","slug":"iclr-2021","status":"publish","type":"msr-event","link":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/event\/iclr-2021\/","title":{"rendered":"Microsoft at ICLR 2021"},"content":{"rendered":"\n\n<p><strong>Conference URL:<\/strong> <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/iclr.cc\/Conferences\/2021\" target=\"_blank\" rel=\"noopener noreferrer\">ICLR 2021<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<p>Microsoft is excited to be a Platinum sponsor of the <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/iclr.cc\/Conferences\/2021\" target=\"_blank\" rel=\"noopener\">Ninth International Conference on Learning Representations (ICLR)<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>. Stop by our virtual booth to chat with our experts, see demos of our latest research and find out more about career opportunities with Microsoft.<\/p>\n<h2>Chat with us!<\/h2>\n<p><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/iclr.cc\/accounts\/login\/?next=\/ExpoConferences\/2021\/Sponsorpage%3Fid%3D95\" target=\"_blank\" rel=\"noopener\">Join us at our virtual booth<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> during the following PST times:<\/p>\n<ul>\n<li>Tuesday, May 4th | 6 AM\u20137 AM, 2 PM\u20133 PM, 10 PM\u201311 PM, 11 PM\u201312 AM<\/li>\n<li>Wednesday, May 5th | 6 AM\u20137 AM, 2 PM\u20133 PM, 10 PM\u201311 PM, 11 PM\u201312 AM<\/li>\n<li>Thursday, May 6th | 6 AM\u20137 AM, 2 PM\u20133 PM, 10 PM\u201311 PM, 11 PM\u201312 AM<\/li>\n<li>Friday, May 7th |\u00a06 AM\u20137 AM, 2 PM\u20133 PM, 10 PM\u201311 PM, 11 PM\u201312 AM<\/li>\n<\/ul>\n<h2>Organizing Committee<\/h2>\n<p><strong>Senior Program Chair:<\/strong> <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/kahofman\/\">Katja Hofmann<\/a><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<h2>Accepted papers<\/h2>\n<p><em>*All times are displayed in PDT<\/em><\/p>\n<h3>Monday, May 3<\/h3>\n<p style=\"margin-bottom: 0px\">01:00 \u2013 03:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>What Makes Instance Discrimination Good for Transfer Learning?<\/strong><\/p>\n<p>Nanxuan Zhao, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/wuzhiron\/\">Zhirong Wu<\/a>, Rynson W. H. Lau, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/stevelin\/\">Stephen Lin<\/a><\/p>\n<p style=\"margin-bottom: 0px\">09:00 \u2013 11:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Learning-based Support Estimation in Sublinear Time<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Talya Eden, Piotr Indyk, Shyam Narayanan, Ronitt Rubinfeld, Sandeep Silwal, <strong>Tal Wagner<\/strong><\/p>\n<p><em>Spotlight Session: May 4<\/em><\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/parameter-efficient-multimodal-transformers-for-video-representation-learning\/\"><strong>Parameter Efficient Multimodal Transformers for Video Representation Learning<\/strong><\/a><\/p>\n<p>Sangho Lee, Youngjae Yu, Gunhee Kim, Thomas Breuel, Jan Kautz, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yalesong\/\">Yale Song<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Shapley Explainability on the Data Manifold<\/strong><\/p>\n<p>Christopher Frye, Damien de Mijolla, Tom Begley, Laurence Cowton, <strong>Megan Stanley<\/strong>, Ilya Feige<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/infobert-improving-robustness-of-language-models-from-an-information-theoretic-perspective\/\"><strong>InfoBERT: Improving Robustness of Language Models from An Information Theoretic Perspective<\/strong><\/a><\/p>\n<p>Boxin Wang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/shuowa\/\">Shuohang Wang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yucheng1\/\">Yu Cheng<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zhgan\/\">Zhe Gan<\/a>, Ruoxi Jia, Bo Li, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jingjl\/\">Jingjing Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\">12:35 \u2013 12:45 PM | Spotlight<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Systematic Generalisation with Group Invariant Predictions<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Faruk Ahmed, Yoshua Bengio, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/havansei\/\">Harm van Seijen<\/a>, Aaron Courville<\/p>\n<p><em>Poster Session: May 4<\/em><\/p>\n<p style=\"margin-bottom: 0px\">05:00 \u2013 07:00 PM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>VA-RED$^2$: Video Adaptive Redundancy Reduction<\/strong><\/p>\n<p>Bowen Pan, Rameswar Panda, Camilo Luciano Fosco, <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/www.linkedin.com\/in\/chung-ching-lin-29b05026\/\" target=\"_blank\" rel=\"noopener\">Chung-Ching Lin<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>, Alex J Andonian, Yue Meng, Kate Saenko, Aude Oliva, Rogerio Feris<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/score-pre-training-for-context-representation-in-conversational-semantic-parsing\/\"><strong>SCoRe: Pre-Training for Context Representation in Conversational Semantic Parsing<\/strong><\/a><\/p>\n<p>Tao Yu, Rui Zhang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/polozov\/\">Alex Polozov<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/meek\/\">Christopher Meek<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/hassanam\/\">Ahmed Hassan Awadallah<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/deberta-decoding-enhanced-bert-with-disentangled-attention-2\/\"><strong>DEBERTA: Decoding-Enhanced BERT with Disentangled Attention<\/strong><\/a><\/p>\n<p><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/penhe\/\">Pengcheng He<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/xiaodl\/\">Xiaodong Liu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jfgao\/\">Jianfeng Gao<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/wzchen\/\">Weizhu Chen<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Learning a Latent Simplex in Input Sparsity Time<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Ainesh Bakshi, Chiranjib Bhattacharyya, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/kannan\/\">Ravi Kannan<\/a>, David Woodruff, Samson Zhou<\/p>\n<p><em>Spotlight Session: May 6<\/em><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Taking Notes on the Fly Helps Language Pre-Training<\/strong><\/p>\n<p>Qiyu Wu, Chen Xing, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yatli\/\">Yatao Li<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/guoke\/\">Guolin Ke<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dihe\/\">Di He<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>MixKD: Towards Efficient Distillation of Large-scale Language Models<\/strong><\/p>\n<p>Kevin J Liang, Weituo Hao, <strong>Dinghan Shen<\/strong>, Yufan Zhou, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/wzchen\/\">Weizhu Chen<\/a>, Changyou Chen, Lawrence Carin<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Rethinking Positional Encoding in Language Pre-training<\/strong><\/p>\n<p><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/guoke\/\">Guolin Ke<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dihe\/\">Di He<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<hr \/>\n<h3>Tuesday, May 4<\/h3>\n<p style=\"margin-bottom: 0px\">01:00 \u2013 03:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/graphcodebert-pre-training-code-representations-with-data-flow\/\"><strong>GraphCodeBERT: Pre-training Code Representations with Data Flow<\/strong><\/a><\/p>\n<p>Daya Guo, Shuo Ren, Shuai Lu, Zhangyin Feng, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dutang\/\">Duyu Tang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/shujliu\/\">Shujie Liu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lozhou\/\">Long Zhou<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/nanduan\/\">Nan Duan<\/a>, <strong>Alexey Svyatkovskiy<\/strong>, <strong>Shengyu Fu<\/strong>, <strong>Michele Tufano<\/strong>, <strong>Shao Kun Deng<\/strong>, <strong>Colin Clement<\/strong>, <strong>Dawn Drain<\/strong>, <strong>Neel Sundaresan<\/strong>, Jian Yin, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/djiang\/\">Daxin Jiang<\/a>, Ming Zhou<\/p>\n<p style=\"margin-bottom: 0px\">09:00 \u2013 11:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/self-supervised-learning-of-compressed-video-representations\/\"><strong>Self-Supervised Learning of Compressed Video Representations<\/strong><\/a><\/p>\n<p>Youngjae Yu, Sangho Lee, Gunhee Kim, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yalesong\/\">Yale Song<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Data-Efficient Reinforcement Learning with Self-Predictive Representations<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Max Schwarzer, <strong>Ankesh Anand<\/strong>, Rishab Goel, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/devonh\/\">R Devon Hjelm<\/a>, Aaron Courville, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/phbachma\/\">Philip Bachman<\/a><\/p>\n<p><em>Spotlight Session: May 6<\/em><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Provable Rich Observation Reinforcement Learning with Combinatorial Latent States<\/strong><\/p>\n<p><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dimisra\/\">Dipendra Misra<\/a>, Qinghua Liu, Chi Jin, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jcl\/\">John Langford<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/approximate-nearest-neighbor-negative-contrastive-learning-for-dense-text-retrieval\/\"><strong>Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval<\/strong><\/a><\/p>\n<p><strong>Lee Xiong<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/cxiong\/\">Chenyan Xiong<\/a>, <strong>Ye Li<\/strong>, <strong>Kwok-Fung Tang<\/strong>, <strong>Jialin Liu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/pauben\/\">Paul N. Bennett<\/a>, <strong>Junaid Ahmed<\/strong>, <strong>Arnold Overwikj<\/strong><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Systematic Generalisation with Group Invariant Predictions<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Faruk Ahmed, Yoshua Bengio, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/havansei\/\">Harm van Seijen<\/a>, Aaron Courville<\/p>\n<p><em>Spotlight Session: May 3<\/em><\/p>\n<p style=\"margin-bottom: 0px\">01:28 \u2013 01:38 PM | Spotlight<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Learning-based Support Estimation in Sublinear Time<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Talya Eden, Piotr Indyk, Shyam Narayanan, Ronitt Rubinfeld, Sandeep Silwal, <strong>Tal Wagner<\/strong><\/p>\n<p><em>Poster Session: May 3<\/em><\/p>\n<p style=\"margin-bottom: 0px\">05:00 \u2013 07:00 PM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>CoDA: Contrast-enhanced and Diversity-promoting Data Augmentation for Natural Language Understanding<\/strong><\/p>\n<p>Yanru Qu, <strong>Dinghan Shen<\/strong>, <strong>Yelong Shen<\/strong>, <strong>Sandra Sajeev<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/wzchen\/\">Weizhu Chen<\/a>, Jiawei Han<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Knowledge Distillation as Semiparametric Inference<\/strong><\/p>\n<p>Tri Dao, <strong>Govinda M Kamath<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/vasy\/\">Vasilis Syrgkanis<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lmackey\/\">Lester Mackey<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Debiasing Concept-based Explanations with Causal Analysis<\/strong><\/p>\n<p>Mohammad Taha Bahadori, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/heckerma\/\">David Heckerman<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Aligning AI with Shared Human Values<\/strong><\/p>\n<p>Dan Hendrycks, Collin Burns, Steven Basart, Andrew Critch, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jerrl\/\">Jerry Li<\/a>, Dawn Song, Jacob Steinhardt<\/p>\n<hr \/>\n<h3>Wednesday, May 5<\/h3>\n<p style=\"margin-bottom: 0px\">01:00 \u2013 03:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/active-contrastive-learning-of-audio-visual-video-representations\/\"><strong>Active Contrastive Learning of Audio-Visual Video Representations<\/strong><\/a><\/p>\n<p><strong>Shuang Ma<\/strong>, Zhaoyang Zeng, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/damcduff\/\">Daniel McDuff<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yalesong\/\">Yale Song<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Dance Revolution: Long-Term Dance Generation with Music via Curriculum Learning<\/strong><\/p>\n<p>Ruozi Huang, <strong>Huang Hu<\/strong>, Wei Wu, Kei Sawada, Mi Zhang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/djiang\/\">Daxin Jiang<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/return-based-contrastive-representation-learning-for-reinforcement-learning\/\"><strong>Return-Based Contrastive Representation Learning for Reinforcement Learning<\/strong><\/a><\/p>\n<p>Guoqing Liu, Chuheng Zhang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lizo\/\">Li Zhao<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a>, Jinhua Zhu, Li Jian, Nenghai Yu, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Byzantine-Resilient Non-Convex Stochastic Gradient Descent<\/strong><\/p>\n<p><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zeyuana\/\">Zeyuan AllenZhu<\/a>, Faeze Ebrahimianghazani, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jerrl\/\">Jerry Li<\/a>, Dan Alistarh<\/p>\n<p style=\"margin-bottom: 0px\">09:00 \u2013 11:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Learning to Represent Action Values as a Hypergraph on the Action Vertices<\/strong><\/p>\n<p>Arash Tavakoli, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/mefatemi\/\">Mehdi Fatemi<\/a>, Petar Kormushev<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Optimism in Reinforcement Learning with Generalized Linear Function Approximation<\/strong><\/p>\n<p>Yining Wang, Ruosong Wang, Simon Shaolei Du, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/akshaykr\/\">Akshay Krishnamurthy<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>SEED: Self-supervised Distillation For Visual Representation<\/strong><\/p>\n<p>Zhiyuan Fang, <strong>Jianfeng Wang<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lijuanw\/\">Lijuan Wang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/leizhang\/\">Lei Zhang<\/a>, Yezhou Yang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zliu\/\">Zicheng Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\">05:00 \u2013 07:00 PM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/filtered-inner-product-projection-for-multilingual-embedding-alignment\/\"><strong>Filtered Inner Product Projection for Crosslingual Embedding Alignment<\/strong><\/a><\/p>\n<p>Vin Sachidananda, Ziyi Yang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/chezhu\/\">Chenguang Zhu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Economic Hyperparameter Optimization With Blended Search Strategy<\/strong><\/p>\n<p><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/chiw\/\">Chi Wang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/qiw\/\">Qingyun Wu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/sihua\/\">Silu Huang<\/a>, <strong>Amin Saied<\/strong><\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/alfworld-aligning-text-and-embodied-environments-for-interactive-learning\/\"><strong>ALFWorld: Aligning Text and Embodied Environments for Interactive Learning<\/strong><\/a><\/p>\n<p>Mohit Shridhar, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/eryua\/\">Xingdi Yuan<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/macote\/\">Marc-Alexandre Cote<\/a>, Yonatan Bisk, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/adtrisch\/\">Adam Trischler<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/mahauskn\/\">Matthew Hausknecht<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>AdaFuse: Adaptive Temporal Fusion Network for Efficient Action Recognition<\/strong><\/p>\n<p>Yue Meng, Rameswar Panda, <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/www.linkedin.com\/in\/chung-ching-lin-29b05026\/\" target=\"_blank\" rel=\"noopener\">Chung-Ching Lin<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>, Prasanna Sattigeri, Leonid Karlinsky, Kate Saenko, Aude Oliva, Rogerio Feris<\/p>\n<p style=\"margin-bottom: 0px\"><strong>AdaSpeech: Adaptive Text to Speech for Custom Voice<\/strong><\/p>\n<p>Mingjian Chen, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/xuta\/\">Xu Tan<\/a>, Bohan Li, Yanqing Liu, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a>, Sheng Zhao, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Revisiting Dynamic Convolution via Matrix Decomposition<\/strong><\/p>\n<p>Yunsheng Li, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yiche\/\">Yinpeng Chen<\/a>, <strong>Xiyang Dai<\/strong>, <strong>Mengchen Liu<\/strong>, <strong>Dongdong Chen<\/strong>, <strong>Ye Yu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/luyuan\/\">Lu Yuan<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zliu\/\">Zicheng Liu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/meic\/\">Mei Chen<\/a>, Nuno Vasconcelos<\/p>\n<p style=\"margin-bottom: 0px\">07:25 \u2013 07:35 PM | Spotlight<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Large Scale Image Completion via Co-Modulated Generative Adversarial Networks<\/strong><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Shengyu Zhao<\/strong>, Jonathan Cui, <strong>Yilun Sheng<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yuedong\/\">Yue Dong<\/a>, Xiao Liang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/echang\/\">Eric I-Chao Chang<\/a>, Yan Xu<\/p>\n<p><em>Poster Session: May 6<\/em><\/p>\n<p style=\"margin-bottom: 0px\">07:55 \u2013 08:10 PM | Oral<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Deformable DETR: Deformable Transformers for End-to-End Object Detection<\/strong><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Xizhou Zhu<\/strong>, <strong>Weijie Su<\/strong>, <strong>Lewei Lu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/libin\/\">Bin Li<\/a>, <strong>Xiaogang Wang<\/strong>, <strong>Jifeng Dai<\/strong><\/p>\n<p><em>Poster Session: May 6<\/em><\/p>\n<hr \/>\n<h3>Thursday, May 6<\/h3>\n<p style=\"margin-bottom: 0px\">01:00 \u2013 03:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/do-not-let-privacy-overbill-utility-gradient-embedding-perturbation-for-private-learning\/\"><strong>Do not Let Privacy Overbill Utility: Gradient Embedding Perturbation for Private Learning<\/strong><\/a><\/p>\n<p><strong>Da Yu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/huzhang\/\">Huishuai Zhang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/weic\/\">Wei Chen<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>IOT: Instance-wise Layer Reordering for Transformer Structures<\/strong><\/p>\n<p>Jinhua Zhu, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lijuwu\/\">Lijun Wu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yinxia\/\">Yingce Xia<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/shufxi\/\">Shufang Xie<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a>, Wengang Zhou, Houqiang Li, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Deformable DETR: Deformable Transformers for End-to-End Object Detection<\/strong><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Xizhou Zhu<\/strong>, <strong>Weijie Su<\/strong>, <strong>Lewei Lu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/libin\/\">Bin Li<\/a>, <strong>Xiaogang Wang<\/strong>, <strong>Jifeng Dai<\/strong><\/p>\n<p><em>Oral Session: May 5<\/em><\/p>\n<p style=\"margin-bottom: 0px\">09:00 \u2013 11:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Initialization and Regularization of Factorized Neural Layers<\/strong><\/p>\n<p>Mikhail Khodak, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/netenenh\/\">Neil A. Tenenholtz<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lmackey\/\">Lester Mackey<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/fusi\/\">Nicolo Fusi<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Learning to Recombine and Resample Data For Compositional Generalization<\/strong><\/p>\n<p>Ekin Aky\u00fcrek, Afra Feyza Aky\u00fcrek, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jaandrea\/\">Jacob Andreas<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/adversarial-score-matching-and-improved-sampling-for-image-generation\/\"><strong>Adversarial score matching and improved sampling for image generation<\/strong><\/a><\/p>\n<p>Alexia Jolicoeur-Martineau, R\u00e9mi Pich\u00e9-Taillefer, Ioannis Mitliagkas, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/retachet\/\">Remi Tachet des Combes<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Improving Zero-Shot Voice Style Transfer via Disentangled Representation Learning<\/strong><\/p>\n<p>Siyang Yuan, Pengyu Cheng, Ruiyi Zhang, Weituo Hao, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zhgan\/\">Zhe Gan<\/a>, Lawrence Carin<\/p>\n<p style=\"margin-bottom: 0px\">12:40 \u2013 12:50 PM | Spotlight<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Data-Efficient Reinforcement Learning with Self-Predictive Representations<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Max Schwarzer, <strong>Ankesh Anand<\/strong>, Rishab Goel, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/devonh\/\">R Devon Hjelm<\/a>, Aaron Courville, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/phbachma\/\">Philip Bachman<\/a><\/p>\n<p><em>Poster Session: May 4<\/em><\/p>\n<p style=\"margin-bottom: 0px\">05:00 \u2013 07:00 PM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Representing Partial Programs with Blended Abstract Semantics<\/strong><\/p>\n<p>Maxwell Nye, Yewen Pu, Matthew Bowers, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jaandrea\/\">Jacob Andreas<\/a>, Joshua B. Tenenbaum, Armando Solar-Lezama<\/p>\n<p style=\"margin-bottom: 0px\"><strong>FastSpeech 2: Fast and High-Quality End-to-End Text to Speech<\/strong><\/p>\n<p>Yi Ren, Chenxu Hu, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/xuta\/\">Xu Tan<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a>, <strong>Sheng Zhao<\/strong>, Zhou Zhao, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Large Scale Image Completion via Co-Modulated Generative Adversarial Networks<\/strong><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Shengyu Zhao<\/strong>, Jonathan Cui, <strong>Yilun Sheng<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yuedong\/\">Yue Dong<\/a>, Xiao Liang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/echang\/\">Eric I-Chao Chang<\/a>, Yan Xu<\/p>\n<p><em>Spotlight Session: May 5<\/em><\/p>\n<p><strong>DynaTune: Dynamic Tensor Program Optimization in Deep Neural Network Compilation<\/strong><br \/>\n<a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/minjiaz\/\">Minjia Zhang<\/a>, <strong>Menghao Li<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/chiw\/\">Chi Wang<\/a>, <strong>Mingqin Li<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">08:58 \u2013 09:08 PM | Spotlight<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Learning a Latent Simplex in Input Sparsity Time<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Ainesh Bakshi, Chiranjib Bhattacharyya, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/kannan\/\">Ravi Kannan<\/a>, David Woodruff, Samson Zhou<\/p>\n<p><em>Poster Session: May 3<\/em><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<h2>Workshops<\/h2>\n<p><em>*All times are displayed in PDT<\/em><\/p>\n<h3>Friday, May 7<\/h3>\n<p style=\"margin-bottom: 0px\">02:30 \u2013 11:30 AM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/sites.google.com\/view\/sedl-workshop\" target=\"_blank\" rel=\"noopener\"><strong>Science and Engineering of Deep Learning<\/strong><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<p>Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dabelgra\/\" rel=\"noopener\">Danielle Belgrave<\/a><\/p>\n<p style=\"margin-bottom: 0px\">05:15 \u2013 10:00 AM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/rethinkingmlpapers.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>Beyond Static Papers: Rethinking How We Share Scientific Understanding in ML<\/strong><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<p>Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/evelynev\/\">Evelyne Viegas<\/a><\/p>\n<p style=\"margin-bottom: 0px\">05:45 AM \u2013 5:45 PM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/sslrlworkshop.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>Self-Supervision for Reinforcement Learning<\/strong><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<p style=\"margin-bottom: 0px\">Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jcl\/\">John Langford<\/a><\/p>\n<p>Co-organizer: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/devonh\/\">R Devon Hjelm<\/a><\/p>\n<p style=\"margin-bottom: 0px\">05:55 AM \u2013 03:00 PM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/aiforpublichealth.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>AI for Public Health<\/strong><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<p>Co-organizer: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/epierson\/\">Emma Pierson<\/a><\/p>\n<p style=\"margin-bottom: 0px\">06:00 AM \u2013 06:00 PM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/iclr2021generalization.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>Generalization beyond the training distribution in brains and machines<\/strong><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<p>Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/idamo\/\">Ida Momennejad<\/a><\/p>\n<p style=\"margin-bottom: 0px\">07:00 AM \u2013 08:30 PM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/sites.google.com\/view\/neucair-workshop\" target=\"_blank\" rel=\"noopener\"><strong>Neural Conversational AI: Bridging the Gap Between Research and Real World (NeuCAIR) Workshop<\/strong><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<p style=\"margin-bottom: 0px\">Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jfgao\/\">Jianfeng Gao<\/a><\/p>\n<p>Co-organizer: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/aslicel\/\">Asli Celikyilmaz<\/a><\/p>\n<p style=\"margin-bottom: 0px\">07:00 AM \u2013 06:30 PM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/sdg-quality-privacy-bias.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>Synthetic Data Generation: Quality, Privacy, Bias<\/strong><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<p>Program Committee Member: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/chezha\/\">Cheng Zhang<\/a><\/p>\n<p style=\"margin-bottom: 0px\">08:00 AM \u2013 05:05 PM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/sites.google.com\/connect.hku.hk\/robustml-2021\/home\" target=\"_blank\" rel=\"noopener\"><strong>Robust and reliable machine learning in the real world<\/strong><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<p style=\"margin-bottom: 0px\">Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/eckamar\/\">Ece Kamar<\/a><\/p>\n<p>Co-organizer: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tristan\/\">Tristan Naumann<\/a><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Microsoft is excited to be a Platinum sponsor of the Ninth International Conference on Learning Representation (ICLR). Stop by our virtual booth to chat with our experts, see demos of our latest research and find out more about career opportunities with Microsoft.<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr_startdate":"2021-05-04","msr_enddate":"2021-05-07","msr_location":"Virtual","msr_expirationdate":"","msr_event_recording_link":"","msr_event_link":"","msr_event_link_redirect":false,"msr_event_time":"","msr_hide_region":false,"msr_private_event":false,"msr_hide_image_in_river":0,"footnotes":""},"research-area":[13556],"msr-region":[256048],"msr-event-type":[197941],"msr-video-type":[],"msr-locale":[268875],"msr-program-audience":[],"msr-post-option":[],"msr-impact-theme":[],"class_list":["post-725710","msr-event","type-msr-event","status-publish","hentry","msr-research-area-artificial-intelligence","msr-region-global","msr-event-type-conferences","msr-locale-en_us"],"msr_about":"<!-- wp:msr\/event-details {\"title\":\"Microsoft at ICLR 2021\",\"backgroundColor\":\"gable-green\",\"imageType\":\"full-bleed\"} \/-->\n\n<!-- wp:msr\/content-tabs --><!-- wp:msr\/content-tab {\"title\":\"About\"} --><!-- wp:freeform --><p><strong>Conference URL:<\/strong> <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/iclr.cc\/Conferences\/2021\" target=\"_blank\" rel=\"noopener noreferrer\">ICLR 2021<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<p>Microsoft is excited to be a Platinum sponsor of the <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/iclr.cc\/Conferences\/2021\" target=\"_blank\" rel=\"noopener\">Ninth International Conference on Learning Representations (ICLR)<\/a>. Stop by our virtual booth to chat with our experts, see demos of our latest research and find out more about career opportunities with Microsoft.<\/p>\n<h2>Chat with us!<\/h2>\n<p><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/iclr.cc\/accounts\/login\/?next=\/ExpoConferences\/2021\/Sponsorpage%3Fid%3D95\" target=\"_blank\" rel=\"noopener\">Join us at our virtual booth<\/a> during the following PST times:<\/p>\n<ul>\n<li>Tuesday, May 4th | 6 AM\u20137 AM, 2 PM\u20133 PM, 10 PM\u201311 PM, 11 PM\u201312 AM<\/li>\n<li>Wednesday, May 5th | 6 AM\u20137 AM, 2 PM\u20133 PM, 10 PM\u201311 PM, 11 PM\u201312 AM<\/li>\n<li>Thursday, May 6th | 6 AM\u20137 AM, 2 PM\u20133 PM, 10 PM\u201311 PM, 11 PM\u201312 AM<\/li>\n<li>Friday, May 7th |\u00a06 AM\u20137 AM, 2 PM\u20133 PM, 10 PM\u201311 PM, 11 PM\u201312 AM<\/li>\n<\/ul>\n<h2>Organizing Committee<\/h2>\n<p><strong>Senior Program Chair:<\/strong> <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/kahofman\/\">Katja Hofmann<\/a><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<!-- \/wp:freeform --><!-- \/wp:msr\/content-tab --><!-- wp:msr\/content-tab {\"title\":\"Accepted papers\"} --><!-- wp:freeform --><h2>Accepted papers<\/h2>\n<p><em>*All times are displayed in PDT<\/em><\/p>\n<h3>Monday, May 3<\/h3>\n<p style=\"margin-bottom: 0px\">01:00 \u2013 03:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>What Makes Instance Discrimination Good for Transfer Learning?<\/strong><\/p>\n<p>Nanxuan Zhao, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/wuzhiron\/\">Zhirong Wu<\/a>, Rynson W. H. Lau, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/stevelin\/\">Stephen Lin<\/a><\/p>\n<p style=\"margin-bottom: 0px\">09:00 \u2013 11:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Learning-based Support Estimation in Sublinear Time<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Talya Eden, Piotr Indyk, Shyam Narayanan, Ronitt Rubinfeld, Sandeep Silwal, <strong>Tal Wagner<\/strong><\/p>\n<p><em>Spotlight Session: May 4<\/em><\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/parameter-efficient-multimodal-transformers-for-video-representation-learning\/\"><strong>Parameter Efficient Multimodal Transformers for Video Representation Learning<\/strong><\/a><\/p>\n<p>Sangho Lee, Youngjae Yu, Gunhee Kim, Thomas Breuel, Jan Kautz, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yalesong\/\">Yale Song<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Shapley Explainability on the Data Manifold<\/strong><\/p>\n<p>Christopher Frye, Damien de Mijolla, Tom Begley, Laurence Cowton, <strong>Megan Stanley<\/strong>, Ilya Feige<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/infobert-improving-robustness-of-language-models-from-an-information-theoretic-perspective\/\"><strong>InfoBERT: Improving Robustness of Language Models from An Information Theoretic Perspective<\/strong><\/a><\/p>\n<p>Boxin Wang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/shuowa\/\">Shuohang Wang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yucheng1\/\">Yu Cheng<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zhgan\/\">Zhe Gan<\/a>, Ruoxi Jia, Bo Li, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jingjl\/\">Jingjing Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\">12:35 \u2013 12:45 PM | Spotlight<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Systematic Generalisation with Group Invariant Predictions<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Faruk Ahmed, Yoshua Bengio, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/havansei\/\">Harm van Seijen<\/a>, Aaron Courville<\/p>\n<p><em>Poster Session: May 4<\/em><\/p>\n<p style=\"margin-bottom: 0px\">05:00 \u2013 07:00 PM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>VA-RED$^2$: Video Adaptive Redundancy Reduction<\/strong><\/p>\n<p>Bowen Pan, Rameswar Panda, Camilo Luciano Fosco, <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/www.linkedin.com\/in\/chung-ching-lin-29b05026\/\" target=\"_blank\" rel=\"noopener\">Chung-Ching Lin<\/a>, Alex J Andonian, Yue Meng, Kate Saenko, Aude Oliva, Rogerio Feris<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/score-pre-training-for-context-representation-in-conversational-semantic-parsing\/\"><strong>SCoRe: Pre-Training for Context Representation in Conversational Semantic Parsing<\/strong><\/a><\/p>\n<p>Tao Yu, Rui Zhang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/polozov\/\">Alex Polozov<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/meek\/\">Christopher Meek<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/hassanam\/\">Ahmed Hassan Awadallah<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/deberta-decoding-enhanced-bert-with-disentangled-attention-2\/\"><strong>DEBERTA: Decoding-Enhanced BERT with Disentangled Attention<\/strong><\/a><\/p>\n<p><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/penhe\/\">Pengcheng He<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/xiaodl\/\">Xiaodong Liu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jfgao\/\">Jianfeng Gao<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/wzchen\/\">Weizhu Chen<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Learning a Latent Simplex in Input Sparsity Time<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Ainesh Bakshi, Chiranjib Bhattacharyya, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/kannan\/\">Ravi Kannan<\/a>, David Woodruff, Samson Zhou<\/p>\n<p><em>Spotlight Session: May 6<\/em><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Taking Notes on the Fly Helps Language Pre-Training<\/strong><\/p>\n<p>Qiyu Wu, Chen Xing, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yatli\/\">Yatao Li<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/guoke\/\">Guolin Ke<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dihe\/\">Di He<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>MixKD: Towards Efficient Distillation of Large-scale Language Models<\/strong><\/p>\n<p>Kevin J Liang, Weituo Hao, <strong>Dinghan Shen<\/strong>, Yufan Zhou, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/wzchen\/\">Weizhu Chen<\/a>, Changyou Chen, Lawrence Carin<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Rethinking Positional Encoding in Language Pre-training<\/strong><\/p>\n<p><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/guoke\/\">Guolin Ke<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dihe\/\">Di He<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<hr \/>\n<h3>Tuesday, May 4<\/h3>\n<p style=\"margin-bottom: 0px\">01:00 \u2013 03:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/graphcodebert-pre-training-code-representations-with-data-flow\/\"><strong>GraphCodeBERT: Pre-training Code Representations with Data Flow<\/strong><\/a><\/p>\n<p>Daya Guo, Shuo Ren, Shuai Lu, Zhangyin Feng, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dutang\/\">Duyu Tang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/shujliu\/\">Shujie Liu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lozhou\/\">Long Zhou<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/nanduan\/\">Nan Duan<\/a>, <strong>Alexey Svyatkovskiy<\/strong>, <strong>Shengyu Fu<\/strong>, <strong>Michele Tufano<\/strong>, <strong>Shao Kun Deng<\/strong>, <strong>Colin Clement<\/strong>, <strong>Dawn Drain<\/strong>, <strong>Neel Sundaresan<\/strong>, Jian Yin, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/djiang\/\">Daxin Jiang<\/a>, Ming Zhou<\/p>\n<p style=\"margin-bottom: 0px\">09:00 \u2013 11:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/self-supervised-learning-of-compressed-video-representations\/\"><strong>Self-Supervised Learning of Compressed Video Representations<\/strong><\/a><\/p>\n<p>Youngjae Yu, Sangho Lee, Gunhee Kim, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yalesong\/\">Yale Song<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Data-Efficient Reinforcement Learning with Self-Predictive Representations<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Max Schwarzer, <strong>Ankesh Anand<\/strong>, Rishab Goel, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/devonh\/\">R Devon Hjelm<\/a>, Aaron Courville, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/phbachma\/\">Philip Bachman<\/a><\/p>\n<p><em>Spotlight Session: May 6<\/em><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Provable Rich Observation Reinforcement Learning with Combinatorial Latent States<\/strong><\/p>\n<p><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dimisra\/\">Dipendra Misra<\/a>, Qinghua Liu, Chi Jin, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jcl\/\">John Langford<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/approximate-nearest-neighbor-negative-contrastive-learning-for-dense-text-retrieval\/\"><strong>Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval<\/strong><\/a><\/p>\n<p><strong>Lee Xiong<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/cxiong\/\">Chenyan Xiong<\/a>, <strong>Ye Li<\/strong>, <strong>Kwok-Fung Tang<\/strong>, <strong>Jialin Liu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/pauben\/\">Paul N. Bennett<\/a>, <strong>Junaid Ahmed<\/strong>, <strong>Arnold Overwikj<\/strong><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Systematic Generalisation with Group Invariant Predictions<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Faruk Ahmed, Yoshua Bengio, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/havansei\/\">Harm van Seijen<\/a>, Aaron Courville<\/p>\n<p><em>Spotlight Session: May 3<\/em><\/p>\n<p style=\"margin-bottom: 0px\">01:28 \u2013 01:38 PM | Spotlight<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Learning-based Support Estimation in Sublinear Time<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Talya Eden, Piotr Indyk, Shyam Narayanan, Ronitt Rubinfeld, Sandeep Silwal, <strong>Tal Wagner<\/strong><\/p>\n<p><em>Poster Session: May 3<\/em><\/p>\n<p style=\"margin-bottom: 0px\">05:00 \u2013 07:00 PM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>CoDA: Contrast-enhanced and Diversity-promoting Data Augmentation for Natural Language Understanding<\/strong><\/p>\n<p>Yanru Qu, <strong>Dinghan Shen<\/strong>, <strong>Yelong Shen<\/strong>, <strong>Sandra Sajeev<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/wzchen\/\">Weizhu Chen<\/a>, Jiawei Han<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Knowledge Distillation as Semiparametric Inference<\/strong><\/p>\n<p>Tri Dao, <strong>Govinda M Kamath<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/vasy\/\">Vasilis Syrgkanis<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lmackey\/\">Lester Mackey<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Debiasing Concept-based Explanations with Causal Analysis<\/strong><\/p>\n<p>Mohammad Taha Bahadori, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/heckerma\/\">David Heckerman<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Aligning AI with Shared Human Values<\/strong><\/p>\n<p>Dan Hendrycks, Collin Burns, Steven Basart, Andrew Critch, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jerrl\/\">Jerry Li<\/a>, Dawn Song, Jacob Steinhardt<\/p>\n<hr \/>\n<h3>Wednesday, May 5<\/h3>\n<p style=\"margin-bottom: 0px\">01:00 \u2013 03:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/active-contrastive-learning-of-audio-visual-video-representations\/\"><strong>Active Contrastive Learning of Audio-Visual Video Representations<\/strong><\/a><\/p>\n<p><strong>Shuang Ma<\/strong>, Zhaoyang Zeng, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/damcduff\/\">Daniel McDuff<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yalesong\/\">Yale Song<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Dance Revolution: Long-Term Dance Generation with Music via Curriculum Learning<\/strong><\/p>\n<p>Ruozi Huang, <strong>Huang Hu<\/strong>, Wei Wu, Kei Sawada, Mi Zhang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/djiang\/\">Daxin Jiang<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/return-based-contrastive-representation-learning-for-reinforcement-learning\/\"><strong>Return-Based Contrastive Representation Learning for Reinforcement Learning<\/strong><\/a><\/p>\n<p>Guoqing Liu, Chuheng Zhang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lizo\/\">Li Zhao<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a>, Jinhua Zhu, Li Jian, Nenghai Yu, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Byzantine-Resilient Non-Convex Stochastic Gradient Descent<\/strong><\/p>\n<p><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zeyuana\/\">Zeyuan AllenZhu<\/a>, Faeze Ebrahimianghazani, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jerrl\/\">Jerry Li<\/a>, Dan Alistarh<\/p>\n<p style=\"margin-bottom: 0px\">09:00 \u2013 11:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Learning to Represent Action Values as a Hypergraph on the Action Vertices<\/strong><\/p>\n<p>Arash Tavakoli, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/mefatemi\/\">Mehdi Fatemi<\/a>, Petar Kormushev<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Optimism in Reinforcement Learning with Generalized Linear Function Approximation<\/strong><\/p>\n<p>Yining Wang, Ruosong Wang, Simon Shaolei Du, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/akshaykr\/\">Akshay Krishnamurthy<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>SEED: Self-supervised Distillation For Visual Representation<\/strong><\/p>\n<p>Zhiyuan Fang, <strong>Jianfeng Wang<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lijuanw\/\">Lijuan Wang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/leizhang\/\">Lei Zhang<\/a>, Yezhou Yang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zliu\/\">Zicheng Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\">05:00 \u2013 07:00 PM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/filtered-inner-product-projection-for-multilingual-embedding-alignment\/\"><strong>Filtered Inner Product Projection for Crosslingual Embedding Alignment<\/strong><\/a><\/p>\n<p>Vin Sachidananda, Ziyi Yang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/chezhu\/\">Chenguang Zhu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Economic Hyperparameter Optimization With Blended Search Strategy<\/strong><\/p>\n<p><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/chiw\/\">Chi Wang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/qiw\/\">Qingyun Wu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/sihua\/\">Silu Huang<\/a>, <strong>Amin Saied<\/strong><\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/alfworld-aligning-text-and-embodied-environments-for-interactive-learning\/\"><strong>ALFWorld: Aligning Text and Embodied Environments for Interactive Learning<\/strong><\/a><\/p>\n<p>Mohit Shridhar, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/eryua\/\">Xingdi Yuan<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/macote\/\">Marc-Alexandre Cote<\/a>, Yonatan Bisk, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/adtrisch\/\">Adam Trischler<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/mahauskn\/\">Matthew Hausknecht<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>AdaFuse: Adaptive Temporal Fusion Network for Efficient Action Recognition<\/strong><\/p>\n<p>Yue Meng, Rameswar Panda, <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/www.linkedin.com\/in\/chung-ching-lin-29b05026\/\" target=\"_blank\" rel=\"noopener\">Chung-Ching Lin<\/a>, Prasanna Sattigeri, Leonid Karlinsky, Kate Saenko, Aude Oliva, Rogerio Feris<\/p>\n<p style=\"margin-bottom: 0px\"><strong>AdaSpeech: Adaptive Text to Speech for Custom Voice<\/strong><\/p>\n<p>Mingjian Chen, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/xuta\/\">Xu Tan<\/a>, Bohan Li, Yanqing Liu, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a>, Sheng Zhao, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Revisiting Dynamic Convolution via Matrix Decomposition<\/strong><\/p>\n<p>Yunsheng Li, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yiche\/\">Yinpeng Chen<\/a>, <strong>Xiyang Dai<\/strong>, <strong>Mengchen Liu<\/strong>, <strong>Dongdong Chen<\/strong>, <strong>Ye Yu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/luyuan\/\">Lu Yuan<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zliu\/\">Zicheng Liu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/meic\/\">Mei Chen<\/a>, Nuno Vasconcelos<\/p>\n<p style=\"margin-bottom: 0px\">07:25 \u2013 07:35 PM | Spotlight<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Large Scale Image Completion via Co-Modulated Generative Adversarial Networks<\/strong><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Shengyu Zhao<\/strong>, Jonathan Cui, <strong>Yilun Sheng<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yuedong\/\">Yue Dong<\/a>, Xiao Liang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/echang\/\">Eric I-Chao Chang<\/a>, Yan Xu<\/p>\n<p><em>Poster Session: May 6<\/em><\/p>\n<p style=\"margin-bottom: 0px\">07:55 \u2013 08:10 PM | Oral<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Deformable DETR: Deformable Transformers for End-to-End Object Detection<\/strong><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Xizhou Zhu<\/strong>, <strong>Weijie Su<\/strong>, <strong>Lewei Lu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/libin\/\">Bin Li<\/a>, <strong>Xiaogang Wang<\/strong>, <strong>Jifeng Dai<\/strong><\/p>\n<p><em>Poster Session: May 6<\/em><\/p>\n<hr \/>\n<h3>Thursday, May 6<\/h3>\n<p style=\"margin-bottom: 0px\">01:00 \u2013 03:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/do-not-let-privacy-overbill-utility-gradient-embedding-perturbation-for-private-learning\/\"><strong>Do not Let Privacy Overbill Utility: Gradient Embedding Perturbation for Private Learning<\/strong><\/a><\/p>\n<p><strong>Da Yu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/huzhang\/\">Huishuai Zhang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/weic\/\">Wei Chen<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>IOT: Instance-wise Layer Reordering for Transformer Structures<\/strong><\/p>\n<p>Jinhua Zhu, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lijuwu\/\">Lijun Wu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yinxia\/\">Yingce Xia<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/shufxi\/\">Shufang Xie<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a>, Wengang Zhou, Houqiang Li, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Deformable DETR: Deformable Transformers for End-to-End Object Detection<\/strong><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Xizhou Zhu<\/strong>, <strong>Weijie Su<\/strong>, <strong>Lewei Lu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/libin\/\">Bin Li<\/a>, <strong>Xiaogang Wang<\/strong>, <strong>Jifeng Dai<\/strong><\/p>\n<p><em>Oral Session: May 5<\/em><\/p>\n<p style=\"margin-bottom: 0px\">09:00 \u2013 11:00 AM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Initialization and Regularization of Factorized Neural Layers<\/strong><\/p>\n<p>Mikhail Khodak, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/netenenh\/\">Neil A. Tenenholtz<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lmackey\/\">Lester Mackey<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/fusi\/\">Nicolo Fusi<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Learning to Recombine and Resample Data For Compositional Generalization<\/strong><\/p>\n<p>Ekin Aky\u00fcrek, Afra Feyza Aky\u00fcrek, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jaandrea\/\">Jacob Andreas<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/adversarial-score-matching-and-improved-sampling-for-image-generation\/\"><strong>Adversarial score matching and improved sampling for image generation<\/strong><\/a><\/p>\n<p>Alexia Jolicoeur-Martineau, R\u00e9mi Pich\u00e9-Taillefer, Ioannis Mitliagkas, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/retachet\/\">Remi Tachet des Combes<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Improving Zero-Shot Voice Style Transfer via Disentangled Representation Learning<\/strong><\/p>\n<p>Siyang Yuan, Pengyu Cheng, Ruiyi Zhang, Weituo Hao, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zhgan\/\">Zhe Gan<\/a>, Lawrence Carin<\/p>\n<p style=\"margin-bottom: 0px\">12:40 \u2013 12:50 PM | Spotlight<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Data-Efficient Reinforcement Learning with Self-Predictive Representations<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Max Schwarzer, <strong>Ankesh Anand<\/strong>, Rishab Goel, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/devonh\/\">R Devon Hjelm<\/a>, Aaron Courville, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/phbachma\/\">Philip Bachman<\/a><\/p>\n<p><em>Poster Session: May 4<\/em><\/p>\n<p style=\"margin-bottom: 0px\">05:00 \u2013 07:00 PM | Poster<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Representing Partial Programs with Blended Abstract Semantics<\/strong><\/p>\n<p>Maxwell Nye, Yewen Pu, Matthew Bowers, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jaandrea\/\">Jacob Andreas<\/a>, Joshua B. Tenenbaum, Armando Solar-Lezama<\/p>\n<p style=\"margin-bottom: 0px\"><strong>FastSpeech 2: Fast and High-Quality End-to-End Text to Speech<\/strong><\/p>\n<p>Yi Ren, Chenxu Hu, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/xuta\/\">Xu Tan<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a>, <strong>Sheng Zhao<\/strong>, Zhou Zhao, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Large Scale Image Completion via Co-Modulated Generative Adversarial Networks<\/strong><\/p>\n<p style=\"margin-bottom: 0px\"><strong>Shengyu Zhao<\/strong>, Jonathan Cui, <strong>Yilun Sheng<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yuedong\/\">Yue Dong<\/a>, Xiao Liang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/echang\/\">Eric I-Chao Chang<\/a>, Yan Xu<\/p>\n<p><em>Spotlight Session: May 5<\/em><\/p>\n<p><strong>DynaTune: Dynamic Tensor Program Optimization in Deep Neural Network Compilation<\/strong><br \/>\n<a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/minjiaz\/\">Minjia Zhang<\/a>, <strong>Menghao Li<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/chiw\/\">Chi Wang<\/a>, <strong>Mingqin Li<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">08:58 \u2013 09:08 PM | Spotlight<\/p>\n<p style=\"margin-bottom: 0px\"><strong>Learning a Latent Simplex in Input Sparsity Time<\/strong><\/p>\n<p style=\"margin-bottom: 0px\">Ainesh Bakshi, Chiranjib Bhattacharyya, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/kannan\/\">Ravi Kannan<\/a>, David Woodruff, Samson Zhou<\/p>\n<p><em>Poster Session: May 3<\/em><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<!-- \/wp:freeform --><!-- \/wp:msr\/content-tab --><!-- wp:msr\/content-tab {\"title\":\"Workshops\"} --><!-- wp:freeform --><h2>Workshops<\/h2>\n<p><em>*All times are displayed in PDT<\/em><\/p>\n<h3>Friday, May 7<\/h3>\n<p style=\"margin-bottom: 0px\">02:30 \u2013 11:30 AM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/sites.google.com\/view\/sedl-workshop\" target=\"_blank\" rel=\"noopener\"><strong>Science and Engineering of Deep Learning<\/strong><\/a><\/p>\n<p>Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dabelgra\/\" rel=\"noopener\">Danielle Belgrave<\/a><\/p>\n<p style=\"margin-bottom: 0px\">05:15 \u2013 10:00 AM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/rethinkingmlpapers.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>Beyond Static Papers: Rethinking How We Share Scientific Understanding in ML<\/strong><\/a><\/p>\n<p>Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/evelynev\/\">Evelyne Viegas<\/a><\/p>\n<p style=\"margin-bottom: 0px\">05:45 AM \u2013 5:45 PM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/sslrlworkshop.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>Self-Supervision for Reinforcement Learning<\/strong><\/a><\/p>\n<p style=\"margin-bottom: 0px\">Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jcl\/\">John Langford<\/a><\/p>\n<p>Co-organizer: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/devonh\/\">R Devon Hjelm<\/a><\/p>\n<p style=\"margin-bottom: 0px\">05:55 AM \u2013 03:00 PM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/aiforpublichealth.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>AI for Public Health<\/strong><\/a><\/p>\n<p>Co-organizer: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/epierson\/\">Emma Pierson<\/a><\/p>\n<p style=\"margin-bottom: 0px\">06:00 AM \u2013 06:00 PM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/iclr2021generalization.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>Generalization beyond the training distribution in brains and machines<\/strong><\/a><\/p>\n<p>Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/idamo\/\">Ida Momennejad<\/a><\/p>\n<p style=\"margin-bottom: 0px\">07:00 AM \u2013 08:30 PM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/sites.google.com\/view\/neucair-workshop\" target=\"_blank\" rel=\"noopener\"><strong>Neural Conversational AI: Bridging the Gap Between Research and Real World (NeuCAIR) Workshop<\/strong><\/a><\/p>\n<p style=\"margin-bottom: 0px\">Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jfgao\/\">Jianfeng Gao<\/a><\/p>\n<p>Co-organizer: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/aslicel\/\">Asli Celikyilmaz<\/a><\/p>\n<p style=\"margin-bottom: 0px\">07:00 AM \u2013 06:30 PM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/sdg-quality-privacy-bias.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>Synthetic Data Generation: Quality, Privacy, Bias<\/strong><\/a><\/p>\n<p>Program Committee Member: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/chezha\/\">Cheng Zhang<\/a><\/p>\n<p style=\"margin-bottom: 0px\">08:00 AM \u2013 05:05 PM | Workshop<\/p>\n<p style=\"margin-bottom: 0px\"><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/sites.google.com\/connect.hku.hk\/robustml-2021\/home\" target=\"_blank\" rel=\"noopener\"><strong>Robust and reliable machine learning in the real world<\/strong><\/a><\/p>\n<p style=\"margin-bottom: 0px\">Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/eckamar\/\">Ece Kamar<\/a><\/p>\n<p>Co-organizer: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tristan\/\">Tristan Naumann<\/a><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<!-- \/wp:freeform --><!-- \/wp:msr\/content-tab --><!-- \/wp:msr\/content-tabs -->","tab-content":[{"id":0,"name":"About","content":"Microsoft is excited to be a Platinum sponsor of the <a href=\"https:\/\/iclr.cc\/Conferences\/2021\" target=\"_blank\" rel=\"noopener\">Ninth International Conference on Learning Representations (ICLR)<\/a>. Stop by our virtual booth to chat with our experts, see demos of our latest research and find out more about career opportunities with Microsoft.\r\n<h2>Chat with us!<\/h2>\r\n<a href=\"https:\/\/iclr.cc\/accounts\/login\/?next=\/ExpoConferences\/2021\/Sponsorpage%3Fid%3D95\" target=\"_blank\" rel=\"noopener\">Join us at our virtual booth<\/a> during the following PST times:\r\n<ul>\r\n \t<li>Tuesday, May 4th | 6 AM\u20137 AM, 2 PM\u20133 PM, 10 PM\u201311 PM, 11 PM\u201312 AM<\/li>\r\n \t<li>Wednesday, May 5th | 6 AM\u20137 AM, 2 PM\u20133 PM, 10 PM\u201311 PM, 11 PM\u201312 AM<\/li>\r\n \t<li>Thursday, May 6th | 6 AM\u20137 AM, 2 PM\u20133 PM, 10 PM\u201311 PM, 11 PM\u201312 AM<\/li>\r\n \t<li>Friday, May 7th |\u00a06 AM\u20137 AM, 2 PM\u20133 PM, 10 PM\u201311 PM, 11 PM\u201312 AM<\/li>\r\n<\/ul>\r\n<h2>Organizing Committee<\/h2>\r\n<strong>Senior Program Chair:<\/strong> <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/kahofman\/\">Katja Hofmann<\/a>"},{"id":1,"name":"Accepted papers","content":"<h2>Accepted papers<\/h2>\r\n<em>*All times are displayed in PDT<\/em>\r\n<h3>Monday, May 3<\/h3>\r\n<p style=\"margin-bottom: 0px\">01:00 \u2013 03:00 AM | Poster<\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>What Makes Instance Discrimination Good for Transfer Learning?<\/strong><\/p>\r\nNanxuan Zhao, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/wuzhiron\/\">Zhirong Wu<\/a>, Rynson W. H. Lau, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/stevelin\/\">Stephen Lin<\/a>\r\n<p style=\"margin-bottom: 0px\">09:00 \u2013 11:00 AM | Poster<\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Learning-based Support Estimation in Sublinear Time<\/strong><\/p>\r\n<p style=\"margin-bottom: 0px\">Talya Eden, Piotr Indyk, Shyam Narayanan, Ronitt Rubinfeld, Sandeep Silwal, <strong>Tal Wagner<\/strong><\/p>\r\n<em>Spotlight Session: May 4<\/em>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/parameter-efficient-multimodal-transformers-for-video-representation-learning\/\"><strong>Parameter Efficient Multimodal Transformers for Video Representation Learning<\/strong><\/a><\/p>\r\nSangho Lee, Youngjae Yu, Gunhee Kim, Thomas Breuel, Jan Kautz, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yalesong\/\">Yale Song<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>Shapley Explainability on the Data Manifold<\/strong><\/p>\r\nChristopher Frye, Damien de Mijolla, Tom Begley, Laurence Cowton, <strong>Megan Stanley<\/strong>, Ilya Feige\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/infobert-improving-robustness-of-language-models-from-an-information-theoretic-perspective\/\"><strong>InfoBERT: Improving Robustness of Language Models from An Information Theoretic Perspective<\/strong><\/a><\/p>\r\nBoxin Wang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/shuowa\/\">Shuohang Wang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yucheng1\/\">Yu Cheng<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zhgan\/\">Zhe Gan<\/a>, Ruoxi Jia, Bo Li, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jingjl\/\">Jingjing Liu<\/a>\r\n<p style=\"margin-bottom: 0px\">12:35 \u2013 12:45 PM | Spotlight<\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Systematic Generalisation with Group Invariant Predictions<\/strong><\/p>\r\n<p style=\"margin-bottom: 0px\">Faruk Ahmed, Yoshua Bengio, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/havansei\/\">Harm van Seijen<\/a>, Aaron Courville<\/p>\r\n<em>Poster Session: May 4<\/em>\r\n<p style=\"margin-bottom: 0px\">05:00 \u2013 07:00 PM | Poster<\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>VA-RED$^2$: Video Adaptive Redundancy Reduction<\/strong><\/p>\r\nBowen Pan, Rameswar Panda, Camilo Luciano Fosco, <a href=\"https:\/\/www.linkedin.com\/in\/chung-ching-lin-29b05026\/\" target=\"_blank\" rel=\"noopener\">Chung-Ching Lin<\/a>, Alex J Andonian, Yue Meng, Kate Saenko, Aude Oliva, Rogerio Feris\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/score-pre-training-for-context-representation-in-conversational-semantic-parsing\/\"><strong>SCoRe: Pre-Training for Context Representation in Conversational Semantic Parsing<\/strong><\/a><\/p>\r\nTao Yu, Rui Zhang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/polozov\/\">Alex Polozov<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/meek\/\">Christopher Meek<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/hassanam\/\">Ahmed Hassan Awadallah<\/a>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/deberta-decoding-enhanced-bert-with-disentangled-attention-2\/\"><strong>DEBERTA: Decoding-Enhanced BERT with Disentangled Attention<\/strong><\/a><\/p>\r\n<a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/penhe\/\">Pengcheng He<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/xiaodl\/\">Xiaodong Liu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jfgao\/\">Jianfeng Gao<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/wzchen\/\">Weizhu Chen<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>Learning a Latent Simplex in Input Sparsity Time<\/strong><\/p>\r\n<p style=\"margin-bottom: 0px\">Ainesh Bakshi, Chiranjib Bhattacharyya, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/kannan\/\">Ravi Kannan<\/a>, David Woodruff, Samson Zhou<\/p>\r\n<em>Spotlight Session: May 6<\/em>\r\n<p style=\"margin-bottom: 0px\"><strong>Taking Notes on the Fly Helps Language Pre-Training<\/strong><\/p>\r\nQiyu Wu, Chen Xing, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yatli\/\">Yatao Li<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/guoke\/\">Guolin Ke<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dihe\/\">Di He<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>MixKD: Towards Efficient Distillation of Large-scale Language Models<\/strong><\/p>\r\nKevin J Liang, Weituo Hao, <strong>Dinghan Shen<\/strong>, Yufan Zhou, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/wzchen\/\">Weizhu Chen<\/a>, Changyou Chen, Lawrence Carin\r\n<p style=\"margin-bottom: 0px\"><strong>Rethinking Positional Encoding in Language Pre-training<\/strong><\/p>\r\n<a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/guoke\/\">Guolin Ke<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dihe\/\">Di He<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a>\r\n\r\n<hr \/>\r\n\r\n<h3>Tuesday, May 4<\/h3>\r\n<p style=\"margin-bottom: 0px\">01:00 \u2013 03:00 AM | Poster<\/p>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/graphcodebert-pre-training-code-representations-with-data-flow\/\"><strong>GraphCodeBERT: Pre-training Code Representations with Data Flow<\/strong><\/a><\/p>\r\nDaya Guo, Shuo Ren, Shuai Lu, Zhangyin Feng, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dutang\/\">Duyu Tang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/shujliu\/\">Shujie Liu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lozhou\/\">Long Zhou<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/nanduan\/\">Nan Duan<\/a>, <strong>Alexey Svyatkovskiy<\/strong>, <strong>Shengyu Fu<\/strong>, <strong>Michele Tufano<\/strong>, <strong>Shao Kun Deng<\/strong>, <strong>Colin Clement<\/strong>, <strong>Dawn Drain<\/strong>, <strong>Neel Sundaresan<\/strong>, Jian Yin, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/djiang\/\">Daxin Jiang<\/a>, Ming Zhou\r\n<p style=\"margin-bottom: 0px\">09:00 \u2013 11:00 AM | Poster<\/p>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/self-supervised-learning-of-compressed-video-representations\/\"><strong>Self-Supervised Learning of Compressed Video Representations<\/strong><\/a><\/p>\r\nYoungjae Yu, Sangho Lee, Gunhee Kim, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yalesong\/\">Yale Song<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>Data-Efficient Reinforcement Learning with Self-Predictive Representations<\/strong><\/p>\r\n<p style=\"margin-bottom: 0px\">Max Schwarzer, <strong>Ankesh Anand<\/strong>, Rishab Goel, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/devonh\/\">R Devon Hjelm<\/a>, Aaron Courville, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/phbachma\/\">Philip Bachman<\/a><\/p>\r\n<em>Spotlight Session: May 6<\/em>\r\n<p style=\"margin-bottom: 0px\"><strong>Provable Rich Observation Reinforcement Learning with Combinatorial Latent States<\/strong><\/p>\r\n<a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dimisra\/\">Dipendra Misra<\/a>, Qinghua Liu, Chi Jin, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jcl\/\">John Langford<\/a>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/approximate-nearest-neighbor-negative-contrastive-learning-for-dense-text-retrieval\/\"><strong>Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval<\/strong><\/a><\/p>\r\n<strong>Lee Xiong<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/cxiong\/\">Chenyan Xiong<\/a>, <strong>Ye Li<\/strong>, <strong>Kwok-Fung Tang<\/strong>, <strong>Jialin Liu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/pauben\/\">Paul N. Bennett<\/a>, <strong>Junaid Ahmed<\/strong>, <strong>Arnold Overwikj<\/strong>\r\n<p style=\"margin-bottom: 0px\"><strong>Systematic Generalisation with Group Invariant Predictions<\/strong><\/p>\r\n<p style=\"margin-bottom: 0px\">Faruk Ahmed, Yoshua Bengio, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/havansei\/\">Harm van Seijen<\/a>, Aaron Courville<\/p>\r\n<em>Spotlight Session: May 3<\/em>\r\n<p style=\"margin-bottom: 0px\">01:28 \u2013 01:38 PM | Spotlight<\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Learning-based Support Estimation in Sublinear Time<\/strong><\/p>\r\n<p style=\"margin-bottom: 0px\">Talya Eden, Piotr Indyk, Shyam Narayanan, Ronitt Rubinfeld, Sandeep Silwal, <strong>Tal Wagner<\/strong><\/p>\r\n<em>Poster Session: May 3<\/em>\r\n<p style=\"margin-bottom: 0px\">05:00 \u2013 07:00 PM | Poster<\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>CoDA: Contrast-enhanced and Diversity-promoting Data Augmentation for Natural Language Understanding<\/strong><\/p>\r\nYanru Qu, <strong>Dinghan Shen<\/strong>, <strong>Yelong Shen<\/strong>, <strong>Sandra Sajeev<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/wzchen\/\">Weizhu Chen<\/a>, Jiawei Han\r\n<p style=\"margin-bottom: 0px\"><strong>Knowledge Distillation as Semiparametric Inference<\/strong><\/p>\r\nTri Dao, <strong>Govinda M Kamath<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/vasy\/\">Vasilis Syrgkanis<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lmackey\/\">Lester Mackey<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>Debiasing Concept-based Explanations with Causal Analysis<\/strong><\/p>\r\nMohammad Taha Bahadori, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/heckerma\/\">David Heckerman<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>Aligning AI with Shared Human Values<\/strong><\/p>\r\nDan Hendrycks, Collin Burns, Steven Basart, Andrew Critch, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jerrl\/\">Jerry Li<\/a>, Dawn Song, Jacob Steinhardt\r\n\r\n<hr \/>\r\n\r\n<h3>Wednesday, May 5<\/h3>\r\n<p style=\"margin-bottom: 0px\">01:00 \u2013 03:00 AM | Poster<\/p>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/active-contrastive-learning-of-audio-visual-video-representations\/\"><strong>Active Contrastive Learning of Audio-Visual Video Representations<\/strong><\/a><\/p>\r\n<strong>Shuang Ma<\/strong>, Zhaoyang Zeng, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/damcduff\/\">Daniel McDuff<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yalesong\/\">Yale Song<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>Dance Revolution: Long-Term Dance Generation with Music via Curriculum Learning<\/strong><\/p>\r\nRuozi Huang, <strong>Huang Hu<\/strong>, Wei Wu, Kei Sawada, Mi Zhang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/djiang\/\">Daxin Jiang<\/a>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/return-based-contrastive-representation-learning-for-reinforcement-learning\/\"><strong>Return-Based Contrastive Representation Learning for Reinforcement Learning<\/strong><\/a><\/p>\r\nGuoqing Liu, Chuheng Zhang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lizo\/\">Li Zhao<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a>, Jinhua Zhu, Li Jian, Nenghai Yu, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>Byzantine-Resilient Non-Convex Stochastic Gradient Descent<\/strong><\/p>\r\n<a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zeyuana\/\">Zeyuan AllenZhu<\/a>, Faeze Ebrahimianghazani, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jerrl\/\">Jerry Li<\/a>, Dan Alistarh\r\n<p style=\"margin-bottom: 0px\">09:00 \u2013 11:00 AM | Poster<\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Learning to Represent Action Values as a Hypergraph on the Action Vertices<\/strong><\/p>\r\nArash Tavakoli, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/mefatemi\/\">Mehdi Fatemi<\/a>, Petar Kormushev\r\n<p style=\"margin-bottom: 0px\"><strong>Optimism in Reinforcement Learning with Generalized Linear Function Approximation<\/strong><\/p>\r\nYining Wang, Ruosong Wang, Simon Shaolei Du, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/akshaykr\/\">Akshay Krishnamurthy<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>SEED: Self-supervised Distillation For Visual Representation<\/strong><\/p>\r\nZhiyuan Fang, <strong>Jianfeng Wang<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lijuanw\/\">Lijuan Wang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/leizhang\/\">Lei Zhang<\/a>, Yezhou Yang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zliu\/\">Zicheng Liu<\/a>\r\n<p style=\"margin-bottom: 0px\">05:00 \u2013 07:00 PM | Poster<\/p>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/filtered-inner-product-projection-for-multilingual-embedding-alignment\/\"><strong>Filtered Inner Product Projection for Crosslingual Embedding Alignment<\/strong><\/a><\/p>\r\nVin Sachidananda, Ziyi Yang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/chezhu\/\">Chenguang Zhu<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>Economic Hyperparameter Optimization With Blended Search Strategy<\/strong><\/p>\r\n<a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/chiw\/\">Chi Wang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/qiw\/\">Qingyun Wu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/sihua\/\">Silu Huang<\/a>, <strong>Amin Saied<\/strong>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/alfworld-aligning-text-and-embodied-environments-for-interactive-learning\/\"><strong>ALFWorld: Aligning Text and Embodied Environments for Interactive Learning<\/strong><\/a><\/p>\r\nMohit Shridhar, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/eryua\/\">Xingdi Yuan<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/macote\/\">Marc-Alexandre Cote<\/a>, Yonatan Bisk, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/adtrisch\/\">Adam Trischler<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/mahauskn\/\">Matthew Hausknecht<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>AdaFuse: Adaptive Temporal Fusion Network for Efficient Action Recognition<\/strong><\/p>\r\nYue Meng, Rameswar Panda, <a href=\"https:\/\/www.linkedin.com\/in\/chung-ching-lin-29b05026\/\" target=\"_blank\" rel=\"noopener\">Chung-Ching Lin<\/a>, Prasanna Sattigeri, Leonid Karlinsky, Kate Saenko, Aude Oliva, Rogerio Feris\r\n<p style=\"margin-bottom: 0px\"><strong>AdaSpeech: Adaptive Text to Speech for Custom Voice<\/strong><\/p>\r\nMingjian Chen, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/xuta\/\">Xu Tan<\/a>, Bohan Li, Yanqing Liu, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a>, Sheng Zhao, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>Revisiting Dynamic Convolution via Matrix Decomposition<\/strong><\/p>\r\nYunsheng Li, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yiche\/\">Yinpeng Chen<\/a>, <strong>Xiyang Dai<\/strong>, <strong>Mengchen Liu<\/strong>, <strong>Dongdong Chen<\/strong>, <strong>Ye Yu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/luyuan\/\">Lu Yuan<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zliu\/\">Zicheng Liu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/meic\/\">Mei Chen<\/a>, Nuno Vasconcelos\r\n<p style=\"margin-bottom: 0px\">07:25 \u2013 07:35 PM | Spotlight<\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Large Scale Image Completion via Co-Modulated Generative Adversarial Networks<\/strong><\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Shengyu Zhao<\/strong>, Jonathan Cui, <strong>Yilun Sheng<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yuedong\/\">Yue Dong<\/a>, Xiao Liang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/echang\/\">Eric I-Chao Chang<\/a>, Yan Xu<\/p>\r\n<em>Poster Session: May 6<\/em>\r\n<p style=\"margin-bottom: 0px\">07:55 \u2013 08:10 PM | Oral<\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Deformable DETR: Deformable Transformers for End-to-End Object Detection<\/strong><\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Xizhou Zhu<\/strong>, <strong>Weijie Su<\/strong>, <strong>Lewei Lu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/libin\/\">Bin Li<\/a>, <strong>Xiaogang Wang<\/strong>, <strong>Jifeng Dai<\/strong><\/p>\r\n<em>Poster Session: May 6<\/em>\r\n\r\n<hr \/>\r\n\r\n<h3>Thursday, May 6<\/h3>\r\n<p style=\"margin-bottom: 0px\">01:00 \u2013 03:00 AM | Poster<\/p>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/do-not-let-privacy-overbill-utility-gradient-embedding-perturbation-for-private-learning\/\"><strong>Do not Let Privacy Overbill Utility: Gradient Embedding Perturbation for Private Learning<\/strong><\/a><\/p>\r\n<strong>Da Yu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/huzhang\/\">Huishuai Zhang<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/weic\/\">Wei Chen<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>IOT: Instance-wise Layer Reordering for Transformer Structures<\/strong><\/p>\r\nJinhua Zhu, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lijuwu\/\">Lijun Wu<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yinxia\/\">Yingce Xia<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/shufxi\/\">Shufang Xie<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a>, Wengang Zhou, Houqiang Li, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>Deformable DETR: Deformable Transformers for End-to-End Object Detection<\/strong><\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Xizhou Zhu<\/strong>, <strong>Weijie Su<\/strong>, <strong>Lewei Lu<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/libin\/\">Bin Li<\/a>, <strong>Xiaogang Wang<\/strong>, <strong>Jifeng Dai<\/strong><\/p>\r\n<em>Oral Session: May 5<\/em>\r\n<p style=\"margin-bottom: 0px\">09:00 \u2013 11:00 AM | Poster<\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Initialization and Regularization of Factorized Neural Layers<\/strong><\/p>\r\nMikhail Khodak, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/netenenh\/\">Neil A. Tenenholtz<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/lmackey\/\">Lester Mackey<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/fusi\/\">Nicolo Fusi<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>Learning to Recombine and Resample Data For Compositional Generalization<\/strong><\/p>\r\nEkin Aky\u00fcrek, Afra Feyza Aky\u00fcrek, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jaandrea\/\">Jacob Andreas<\/a>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/publication\/adversarial-score-matching-and-improved-sampling-for-image-generation\/\"><strong>Adversarial score matching and improved sampling for image generation<\/strong><\/a><\/p>\r\nAlexia Jolicoeur-Martineau, R\u00e9mi Pich\u00e9-Taillefer, Ioannis Mitliagkas, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/retachet\/\">Remi Tachet des Combes<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>Improving Zero-Shot Voice Style Transfer via Disentangled Representation Learning<\/strong><\/p>\r\nSiyang Yuan, Pengyu Cheng, Ruiyi Zhang, Weituo Hao, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/zhgan\/\">Zhe Gan<\/a>, Lawrence Carin\r\n<p style=\"margin-bottom: 0px\">12:40 \u2013 12:50 PM | Spotlight<\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Data-Efficient Reinforcement Learning with Self-Predictive Representations<\/strong><\/p>\r\n<p style=\"margin-bottom: 0px\">Max Schwarzer, <strong>Ankesh Anand<\/strong>, Rishab Goel, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/devonh\/\">R Devon Hjelm<\/a>, Aaron Courville, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/phbachma\/\">Philip Bachman<\/a><\/p>\r\n<em>Poster Session: May 4<\/em>\r\n<p style=\"margin-bottom: 0px\">05:00 \u2013 07:00 PM | Poster<\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Representing Partial Programs with Blended Abstract Semantics<\/strong><\/p>\r\nMaxwell Nye, Yewen Pu, Matthew Bowers, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jaandrea\/\">Jacob Andreas<\/a>, Joshua B. Tenenbaum, Armando Solar-Lezama\r\n<p style=\"margin-bottom: 0px\"><strong>FastSpeech 2: Fast and High-Quality End-to-End Text to Speech<\/strong><\/p>\r\nYi Ren, Chenxu Hu, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/xuta\/\">Xu Tan<\/a>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a>, <strong>Sheng Zhao<\/strong>, Zhou Zhao, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a>\r\n<p style=\"margin-bottom: 0px\"><strong>Large Scale Image Completion via Co-Modulated Generative Adversarial Networks<\/strong><\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Shengyu Zhao<\/strong>, Jonathan Cui, <strong>Yilun Sheng<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/yuedong\/\">Yue Dong<\/a>, Xiao Liang, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/echang\/\">Eric I-Chao Chang<\/a>, Yan Xu<\/p>\r\n<em>Spotlight Session: May 5<\/em>\r\n\r\n<strong>DynaTune: Dynamic Tensor Program Optimization in Deep Neural Network Compilation<\/strong>\r\n<a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/minjiaz\/\">Minjia Zhang<\/a>, <strong>Menghao Li<\/strong>, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/chiw\/\">Chi Wang<\/a>, <strong>Mingqin Li<\/strong>\r\n<p style=\"margin-bottom: 0px\">08:58 \u2013 09:08 PM | Spotlight<\/p>\r\n<p style=\"margin-bottom: 0px\"><strong>Learning a Latent Simplex in Input Sparsity Time<\/strong><\/p>\r\n<p style=\"margin-bottom: 0px\">Ainesh Bakshi, Chiranjib Bhattacharyya, <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/kannan\/\">Ravi Kannan<\/a>, David Woodruff, Samson Zhou<\/p>\r\n<em>Poster Session: May 3<\/em>"},{"id":2,"name":"Workshops","content":"<h2>Workshops<\/h2>\r\n<em>*All times are displayed in PDT<\/em>\r\n<h3>Friday, May 7<\/h3>\r\n<p style=\"margin-bottom: 0px\">02:30 \u2013 11:30 AM | Workshop<\/p>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/sites.google.com\/view\/sedl-workshop\" target=\"_blank\" rel=\"noopener\"><strong>Science and Engineering of Deep Learning<\/strong><\/a><\/p>\r\nSpeaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/dabelgra\/\" rel=\"noopener\">Danielle Belgrave<\/a>\r\n<p style=\"margin-bottom: 0px\">05:15 \u2013 10:00 AM | Workshop<\/p>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/rethinkingmlpapers.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>Beyond Static Papers: Rethinking How We Share Scientific Understanding in ML<\/strong><\/a><\/p>\r\nSpeaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/evelynev\/\">Evelyne Viegas<\/a>\r\n<p style=\"margin-bottom: 0px\">05:45 AM \u2013 5:45 PM | Workshop<\/p>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/sslrlworkshop.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>Self-Supervision for Reinforcement Learning<\/strong><\/a><\/p>\r\n<p style=\"margin-bottom: 0px\">Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jcl\/\">John Langford<\/a><\/p>\r\nCo-organizer: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/devonh\/\">R Devon Hjelm<\/a>\r\n<p style=\"margin-bottom: 0px\">05:55 AM \u2013 03:00 PM | Workshop<\/p>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/aiforpublichealth.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>AI for Public Health<\/strong><\/a><\/p>\r\nCo-organizer: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/epierson\/\">Emma Pierson<\/a>\r\n<p style=\"margin-bottom: 0px\">06:00 AM \u2013 06:00 PM | Workshop<\/p>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/iclr2021generalization.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>Generalization beyond the training distribution in brains and machines<\/strong><\/a><\/p>\r\nSpeaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/idamo\/\">Ida Momennejad<\/a>\r\n<p style=\"margin-bottom: 0px\">07:00 AM \u2013 08:30 PM | Workshop<\/p>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/sites.google.com\/view\/neucair-workshop\" target=\"_blank\" rel=\"noopener\"><strong>Neural Conversational AI: Bridging the Gap Between Research and Real World (NeuCAIR) Workshop<\/strong><\/a><\/p>\r\n<p style=\"margin-bottom: 0px\">Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/jfgao\/\">Jianfeng Gao<\/a><\/p>\r\nCo-organizer: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/aslicel\/\">Asli Celikyilmaz<\/a>\r\n<p style=\"margin-bottom: 0px\">07:00 AM \u2013 06:30 PM | Workshop<\/p>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/sdg-quality-privacy-bias.github.io\/\" target=\"_blank\" rel=\"noopener\"><strong>Synthetic Data Generation: Quality, Privacy, Bias<\/strong><\/a><\/p>\r\nProgram Committee Member: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/chezha\/\">Cheng Zhang<\/a>\r\n<p style=\"margin-bottom: 0px\">08:00 AM \u2013 05:05 PM | Workshop<\/p>\r\n<p style=\"margin-bottom: 0px\"><a href=\"https:\/\/sites.google.com\/connect.hku.hk\/robustml-2021\/home\" target=\"_blank\" rel=\"noopener\"><strong>Robust and reliable machine learning in the real world<\/strong><\/a><\/p>\r\n<p style=\"margin-bottom: 0px\">Speaker: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/eckamar\/\">Ece Kamar<\/a><\/p>\r\nCo-organizer: <a href=\"https:\/\/newed.any0.dpdns.org\/en-us\/research\/people\/tristan\/\">Tristan Naumann<\/a>"}],"msr_startdate":"2021-05-04","msr_enddate":"2021-05-07","msr_event_time":"","msr_location":"Virtual","msr_event_link":"","msr_event_recording_link":"","msr_startdate_formatted":"May 4, 2021","msr_register_text":"Watch now","msr_cta_link":"","msr_cta_text":"","msr_cta_bi_name":"","featured_image_thumbnail":null,"event_excerpt":"Microsoft is excited to be a Platinum sponsor of the Ninth International Conference on Learning Representation (ICLR). Stop by our virtual booth to chat with our experts, see demos of our latest research and find out more about career opportunities with Microsoft.","msr_research_lab":[],"related-researchers":[],"msr_impact_theme":[],"related-academic-programs":[],"related-groups":[],"related-projects":[],"related-opportunities":[],"related-publications":[671469,732229,742414,787795],"related-videos":[],"related-posts":[734179,742696,746077],"_links":{"self":[{"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/725710","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-event"}],"about":[{"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-event"}],"version-history":[{"count":5,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/725710\/revisions"}],"predecessor-version":[{"id":1146895,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/725710\/revisions\/1146895"}],"wp:attachment":[{"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/media?parent=725710"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=725710"},{"taxonomy":"msr-region","embeddable":true,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-region?post=725710"},{"taxonomy":"msr-event-type","embeddable":true,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-event-type?post=725710"},{"taxonomy":"msr-video-type","embeddable":true,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-video-type?post=725710"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=725710"},{"taxonomy":"msr-program-audience","embeddable":true,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-program-audience?post=725710"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=725710"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/newed.any0.dpdns.org\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=725710"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}