ICLR 2018入选论文公布：23篇Oral，508篇被拒

岳排槐 发自 凹非寺
量子位 出品 | 公众号 QbitAI

好消息，今年的ICLR入选论文，今早正式公布。

ICLR是深度学习领域的顶级会议，由深度学习三巨头中的Yoshua Bengio和Yann LeCun主导创办。他们二人仍然是今年ICLR的大会主席。

将于4月底召开的第六届ICLR总共收到了900多篇论文，同比增加一倍。这个大会采用公开评审制度，所有论文会匿名公开在open review网站上，接受同行们的匿名评分和提问。

此前有博客统计了今年ICLR论文的评分情况，平均分是5.24，中位数是5.33。

最新公布的数据显示，ICLR 2018共有23篇论文入围Oral Papers，另有314篇论文获评Poster Papers。另有90篇论文被选为Workshop Papers，508篇论文被拒，63篇论文撤回。

入围的Oral Papers中，此前评分在8.0以上的包括：

• Certifiable Distributional Robustness with Principled Adversarial Training

• On the Convergence of Adam and Beyond

• Deep Mean Field Games for Learning Optimal Behavior Policy of Large Populations

• Multi-Scale Dense Networks for Resource Efficient Image Classification

• Boosting Dilated Convolutional Networks with Mixed Tensor Decompositions

• Learning to Represent Programs with Graphs

• Wasserstein Auto-Encoders

此前评分较高，但没有中Oral的论文包括：

• Emergence of grid-like representations by training recurrent neural networks to perform spatial localization

• i-RevNet: Deep Invertible Networks

• Stabilizing Adversarial Nets with Prediction Methods

不过这三篇都在Poster Papers之列。

以下是入围Oral Papers的全部论文：

• AmbientGAN: Generative models from lossy measurements

  Ashish Bora, Eric Price, Alexandros G. Dimakis

  https://openreview.net/forum?id=Hy7fDog0b

  评分：7.33

• Beyond Word Importance: Contextual Decomposition to Extract Interactions from LSTMs

  W. James Murdoch, Peter J. Liu, Bin Yu

  https://openreview.net/forum?id=rkRwGg-0Z

  评分：5.56

• Zero-Shot Visual Imitation

  Deepak Pathak, Parsa Mahmoudieh, Michael Luo, Pulkit Agrawal, Dian Chen, Fred Shentu, Evan Shelhamer, Jitendra Malik, Alexei A. Efros, Trevor Darrell

  https://openreview.net/forum?id=BkisuzWRW

  评分：6.91

• Variance Reduction for Policy Gradient with Action-Dependent Factorized Baselines

  Cathy Wu, Aravind Rajeswaran, Yan Duan, Vikash Kumar, Alexandre M Bayen, Sham Kakade, Igor Mordatch, Pieter Abbeel

  https://openreview.net/forum?id=H1tSsb-AW

  评分：6.91

• Progressive Growing of GANs for Improved Quality, Stability, and Variation

  Tero Karras, Timo Aila, Samuli Laine, Jaakko Lehtinen

  https://openreview.net/forum?id=Hk99zCeAb

  评分：5.67

• Neural Sketch Learning for Conditional Program Generation

  Vijayaraghavan Murali, Letao Qi, Swarat Chaudhuri, Chris Jermaine

  https://openreview.net/forum?id=HkfXMz-Ab

  评分：7.44

• Boosting Dilated Convolutional Networks with Mixed Tensor Decompositions

  Nadav Cohen, Ronen Tamari, Amnon Shashua

  https://openreview.net/forum?id=S1JHhv6TW

  评分：8.00

• Continuous Adaptation via Meta-Learning in Nonstationary and Competitive Environments

  Maruan Al-Shedivat, Trapit Bansal, Yura Burda, Ilya Sutskever, Igor Mordatch, Pieter Abbeel

  https://openreview.net/forum?id=Sk2u1g-0-

  评分：7.80

• Breaking the Softmax Bottleneck: A High-Rank RNN Language Model

  Zhilin Yang, Zihang Dai, Ruslan Salakhutdinov, William W. Cohen

  https://openreview.net/forum?id=HkwZSG-CZ

  评分：6.31

• Characterizing Adversarial Subspaces Using Local Intrinsic Dimensionality

  Xingjun Ma, Bo Li, Yisen Wang, Sarah M. Erfani, Sudanthi Wijewickrema, Grant Schoenebeck, Michael E. Houle, Dawn Song, James Bailey

  https://openreview.net/forum?id=B1gJ1L2aW

  评分：7.25

• Learning to Represent Programs with Graphs

  Miltiadis Allamanis, Marc Brockschmidt, Mahmoud Khademi

  https://openreview.net/forum?id=BJOFETxR-

  评分：8.00

• Spectral Normalization for Generative Adversarial Networks

  Takeru Miyato, Toshiki Kataoka, Masanori Koyama, Yuichi Yoshida

  https://openreview.net/forum?id=B1QRgziT-

  评分：7.33

• Wasserstein Auto-Encoders

  Ilya Tolstikhin, Olivier Bousquet, Sylvain Gelly, Bernhard Schoelkopf

  https://openreview.net/forum?id=HkL7n1-0b

  评分：8.00

• Deep Mean Field Games for Learning Optimal Behavior Policy of Large Populations

  Jiachen Yang, Xiaojing Ye, Rakshit Trivedi, Huan Xu, Hongyuan Zha

  https://openreview.net/forum?id=HktK4BeCZ

  评分：8.50

• Certifiable Distributional Robustness with Principled Adversarial Training

  Aman Sinha, Hongseok Namkoong, John Duchi

  https://openreview.net/forum?id=Hk6kPgZA-

  评分：9.00

• On the insufficiency of existing momentum schemes for Stochastic Optimization

  Rahul Kidambi, Praneeth Netrapalli, Prateek Jain, Sham M. Kakade

  https://openreview.net/forum?id=rJTutzbA-

  评分：6.14

• Ask the Right Questions: Active Question Reformulation with Reinforcement Learning

  Christian Buck, Jannis Bulian, Massimiliano Ciaramita, Wojciech Gajewski, Andrea Gesmundo, Neil Houlsby, Wei Wang.

  https://openreview.net/forum?id=S1CChZ-CZ

  评分：5.83

• Spherical CNNs

  Taco S. Cohen, Mario Geiger, Jonas Köhler, Max Welling

  https://openreview.net/forum?id=Hkbd5xZRb

  评分：7.57

• Emergence of Linguistic Communication from Referential Games with Symbolic and Pixel Input

  Angeliki Lazaridou, Karl Moritz Hermann, Karl Tuyls, Stephen Clark

  https://openreview.net/forum?id=HJGv1Z-AW

  评分：6.85

• Training and Inference with Integers in Deep Neural Networks

  Shuang Wu, Guoqi Li, Luping Shi, Feng Chen

  https://openreview.net/forum?id=HJGXzmspb

  评分：6.27

• Multi-Scale Dense Networks for Resource Efficient Image Classification

  Gao Huang, Danlu Chen, Tianhong Li, Felix Wu, Laurens van der Maaten, Kilian Weinberger

  https://openreview.net/forum?id=Hk2aImxAb

  评分：8.33

• Synthetic and Natural Noise Both Break Neural Machine Translation

  Yonatan Belinkov, Yonatan Bisk

  https://openreview.net/forum?id=BJ8vJebC-

  评分：7.33

• On the Convergence of Adam and Beyond

  Sashank J. Reddi, Satyen Kale, Sanjiv Kumar

  https://openreview.net/forum?id=ryQu7f-RZ

  评分：8.56

OMT

量子位还制作了一份《2018年人工智能顶会月历》，也列在今天的推送中，欢迎有需要的朋友们自行取用。

— 完 —

活动报名

加入社群

量子位AI社群13群开始招募啦，欢迎对AI感兴趣的同学，加小助手微信qbitbot4入群；

此外，量子位专业细分群(自动驾驶、CV、NLP、机器学习等)正在招募，面向正在从事相关领域的工程师及研究人员。

进群请加小助手微信号qbitbot5，并务必备注相应群的关键词~通过审核后我们将邀请进群。（专业群审核较严，敬请谅解）

诚挚招聘

量子位正在招募编辑/记者，工作地点在北京中关村。期待有才气、有热情的同学加入我们！相关细节，请在量子位公众号(QbitAI)对话界面，回复“招聘”两个字。

量子位 QbitAI · 头条号签约作者

վ'ᴗ' ի 追踪AI技术和产品新动态