.svg)
Are you ready for NeurIPS 2018?
We mentioned in a previous post how excited we are that eight papers by Petuum authors were accepted to NeurIPS 2018, the 32nd Conference on Neural Information Processing Systems. They’ll be presented at the event next week, which runs from December 3–8.
Although we are a relatively young, mid-sized startup, Petuum is among the top eight corporate research institutions with the most accepted papers this year, along with Google Research, Microsoft, Deepmind, Facebook, and IBM Research. This achievement speaks to the exceptional quality of our research team and the cutting-edge nature of the area we are exploring.
Additionally, we are happy to share the news that Petuum placed in the top three entries for each track of the NeurIPS Adversarial Vision Challenge 2018, winning two of them. Our team, Petuum-CMU, placed first in both the Robust Model Track and Targeted Attack Track, and third in the Untargeted Attack Track. The team included Yaodong Yu, our intern, and two of our summer interns Hongyang Zhang and Susu Xu, in addition to Hongbao Zhang (a Senior Machine Learning Engineer at Petuum), Pengtao Xie (our Senior Director of Data Solutions and Service), and Eric Xing (Petuum’s founder, CEO, and Chief Scientist). If you are interested in hearing about how we approached the challenge, the team will be presenting their strategies along with the other winners during the competition workshop on Friday, December 7, at 9:00 AM in Room 518 as part of the NeurIPS 2018 Competition Track Day 1. See the rest of the results here.
If you would like to catch any of our spotlights, paper presentations, or other workshops, here’s the line-up. For those of you who can’t make it, we will be publishing blog posts describing each of these papers in the coming months. The first one, on the development of a hybrid retrieval-generation reinforced agent (HRGR-Agent), is already up! Read it here.
Spotlights
Tuesday Track 3 Session 2: December 4, 4:45-4:50 PM in Room 517 CD
- #166 Neural Architecture Search with Bayesian Optimisation and Optimal Transport, by Kirthevasan Kandasamy, Willie Neiswanger, Jeff Schneider, Barnabas Poczos, and Eric Xing
Wednesday Track 2 Session 2: December 5, 4:40-4:45 PM in Room 220 E
- #9 DAGs with NO TEARS: Continuous Optimization for Structure Learning, by Xun Zheng, Bryon Aragam, Pradeep Ravikumar, and Eric Xing
Posters
Tuesday Poster Session A: December 4, 10:45 AM-12:45 PM in Room 210 & 230 AB
- #155 Unsupervised Text Style Transfer using Language Models as Discriminators, by Zichao Yang, Zhiting Hu, Chris Dyer, Eric Xing, and Taylor Berg-Kirkpatrick
Tuesday Poster Session B: December 4, 5-7 PM in Room 210 & 230 AB
- #56 Hybrid Retrieval-Generation Reinforced Agent for Medical Image Report Generation, by Christy Y. Li, Xiaodan Liang, Zhiting Hu, and Eric Xing
- #166 Neural Architecture Search with Bayesian Optimisation and Optimal Transport, by Kirthevasan Kandasamy, Willie Neiswanger, Jeff Schneider, Barnabas Poczos, and Eric Xing
Wednesday Poster Session A: December 5, 10:45 AM-12:45 PM in Room 210 & 230 AB
- #32 Deep Generative Models with Learnable Knowledge Constraints, by Zhiting Hu, Zichao Yang, Ruslan Salakhutdinov, Xiaodan Liang, Lianhui Qin, Haoye Dong, and Eric Xing
- #131 Learning Pipelines with Limited Data and Domain Knowledge: A Study in Parsing Physics Problems, by Mrinmaya Sachan, Avinava Dubey, Tom Mitchell, Dan Roth, and Eric Xing
Wednesday Poster Session B: December 5, 5-7 PM in Room 210 & 230 AB
- #9 DAGs with NO TEARS: Continuous Optimization for Structure Learning, by Xun Zheng, Bryon Aragam, Pradeep Ravikumar, and Eric Xing
Thursday Poster Session A: December 6, 10:45 AM-12:45 PM in Room 210 & 230 AB
- #117 Symbolic Graph Reasoning Meets Convolutions, by Xiaodan Liang, Zhiting Hu, and Eric Xing
Thursday Poster Session B: December 6, 5–7 PM in Room 210 & 230 AB
- #96 The Sample Complexity of Semi-Supervised Learning with Nonparametric Mixture Models, by Chen Dan, Leqi Liu, Bryon Aragam, Pradeep Ravikumar, and Eric Xing
Workshops
Friday, December 7, 2:30 PM in Room 510 ABCD: MLSys: Workshop on Systems for ML and Open Source Software
- Dynamic Scheduling For Dynamic Control Flow in Deep Learning Systems, by Jinliang Wei, Vijay Vasudevan, Garth Gibson, and Eric Xing
Saturday, December 8, 9:45 AM in Room 517D: Machine Learning for Health (ML4H): Moving Beyond supervised learning in healthcare
- Unsupervised Pseudo-Labeling for Extractive Summarization on Electronic Health Records, by Xiangan Liu, Keyang Xu, Pengtao Xie, and Eric Xing
We cannot emphasize enough how proud we are of our researchers. If you would like to join us for a celebration after the event in honor of the hard work our team members have done, please let us know by registering via this form.
See you there!
