John Hewitt
Hello! I’m a second-year PhD student in computer science, conducting research in natural language processing at Stanford University. I am grateful to be co-advised by Chris Manning and Percy Liang.
I design and analyze machine learning models of human languages. I’m broadly interested in (interpreting) representation learning, induction of latent hierarchical structure, settings with small data, and multilinguality.
As an undergraduate at Penn, I worked in the lab of Chris Callison-Burch.
Feel free to look me up on Google Scholar or Twitter, or take my CV.
News
- [November 2019] I’m giving a talk at Berkeley on Probing Neural NLP: Ideas and Problems!
- [November 2019] Designing and Interpreting Probes with Control Tasks was named best paper runner up at EMNLP 2019! Slides of the talk available here.
- [August 2019] My work with Percy Liang on designing and understanding neural probes with random control tasks has been accepted to EMNLP! paper and blog post are both available.
- [July 2019] I’m giving a talk at Amazon AI on finding and understanding emergent linguistic structure in neural NLP!
- [June 2019] Gave a talk at NAACL’19 on structural probes! PDF of the slides now available.
- [April 2019] I had a great time being interviewed by Waleed Ammar and Matt Gardner on structural probes! You can find the new NLP Highlights podcast episode on the topic here.
- [March 2019] I’m presenting a poster on syntax in unsupervised representations of language at the Stanford Human-Centered Artificial Intelligence Institute Symposium
- [Feb 2019] My work with Chris Manning on methods for finding syntax trees embedded in contextual representations of language has been accepted to NAACL 2019!
- [Oct 2018] I’ve started offering office hours for research-interested Stanford undergraduates!
Research Office Hours for Undergraduates (ROHU)
An open time for undergraduates looking for advice and discussions on natural language processing research. learn more.
- When Wednesdays, 6:00-7:00 PM (starting Oct 9!)
- Where Gates Building, 2nd floor, A-wing big middle room (219). (Or 219a, in the back, if 219 is busy.)
- Who You, Stanford undergraduates wanting to chat about research; me, wanting to help
-
-
Publications
Publications
2019
Designing and Interpreting Probes with Control Tasks.
John Hewitt and Percy Liang.
EMNLP 2019 (long papers). Runner up best paper. (pdf) (blog) (code) (codalab) (slides).
A Structual Probe for Finding Syntax in Word Representations.
John Hewitt and Christopher D. Manning.
NAACL 2019 (short papers). (pdf) (bib) (blog) (code) (nlp highlights podcast) (slides). (talk)
Simple, Fast, Accurate Intent Classification and Slot Labeling for Goal-Oriented Dialogue Systems.
Arshit Gupta*, John Hewitt* and Katrin Kirchhoff.
SIGDIAL 2019. (pdf)
*: Equal contribution; authors listed alphabetically2018
A Distributional and Orthographic Aggregation Model for English Derivational Morphology.
Daniel Deutsch*, John Hewitt* and Dan Roth.
ACL 2018 (long papers). (pdf)
*: Equal contribution; authors listed alphabeticallyLearning Translations via Images with a Massively Multilingual Image Dataset.
John Hewitt*, Daphne Ippolito*, Brendan Callahan, Reno Kriz, Derry Tanti Wijaya and Chris Callison-Burch.
ACL 2018 (long papers). (pdf)
*: Equal contribution; authors listed alphabeticallyXNMT: The eXtensible Neural Machine Translation Toolkit.
Graham Neubig, Matthias Sperber, Xinyi Wang, Matthieu Felix, Austin Matthews, Sarguna Padmanabhan, Ye Qi, Devendra Singh Sachan, Philip Arthur, Pierre Godard, John Hewitt, Rachid Riad, and Liming Wang.
AMTA 2018. (pdf)2017
Learning Translations via Matrix Completion.
Derry Tanti Wijaya, Brendan Callahan, John Hewitt , Xiao Ling, Marianna Apidianaki, and Chris Callison-Burch.
EMNLP 2017 (long papers) (pdf)Semantic Bootstrapping in Frames: A Computational Model of Syntactic Category Acquisition.
John Hewitt, Jordan Kodner, Mitch Marcus, and Charles Yang.
Conference of the Cognitive Science Society (CogSci), (member posters) 2017. (pdf) (abstract)2016
Automatic Construction of Morphologically-Motivated Translation Models for Highly Inflected Low-Resource Languages.
John Hewitt, Matt Post, David Yarowsky.
AMTA 2016 (pdf)Research Talks
Emergent Linguistic Structure in Neural NLP.
Amazon AI. July 25, 2019.A Distributional and Orthographic Aggregation Model for English Derivational Morphology.
Stanford NLP meeting. Nov 8, 2018.Phrasal Translation Generation for Highly Inflected, Low-Resource Languages.
Penn Computational Linguistics and Lunch (CLunch). Oct 28, 2015.Patents
Capturing Rich Response Relationships with Small-Data Neural Networks.
John Hewitt.
US Patent App 15/841,963. December 2017. (granted). (application) - Blog
-
Projects
-
seq2seq lecture
- I wrote a lecture on sequence-to-sequence models, focusing on visual intuition and assuming very little student background, for Penn’s computational linguistics class with Reno Kriz. Check it out and feel free to give comments.
-
Massively Multilingual Image Dataset
- This large, massively multilingual dataset of images parallel to the words they represent, those words’ translation into English, and images of those English words, is described in an ACL 2018 paper.
-
-
About
Tidbits
Scott Aaronson’s old note on frameworks for reasoning about large numbers, for enjoyment
Kevin Knight’s note on unix commands, to help you with your bash skills
The Fundamental Whiteboard Difficulty (Scott Aaronson):
I figured that chalk has its problems—it breaks, the dust gets all over—but I could live with them, much more than I could live with the Fundamental Whiteboard Difficulty, of all the available markers always being dry whenever you want to explain anything.
I highly suggest Arch Linux for its configurability and the educational experience it provides…
Contact
Take my school email johnhew@stanford, and predict the TLD using your internal knowledge base.