Particularly love the appreciation of "simple" but effective methods! 💕

Our work on learned storage systems and query optimizers is featured on MIT News. Super proud on my team and collaborators, especially Jialin Ding and Ryan Marcus who lead the work news.mit.edu/2020/mit-data-…

Introducing an early-release version of the Language Interpretability Tool (LIT), a visual, interactive, and extensible open-source tool for analyzing all sorts of NLP models 🔥 Code: github.com/pair-code/lit/ Paper: arxiv.org/abs/2008.05122 #NLProc (1/4)

Excited to announce TensorFlow Recommenders, just in time for #RecSys2020!

Super happy to share that our paper “Fairness without Demographics through Adversarially Reweighted Learning" has been accepted at #NeurIPS2020. A big thanks to all my co-authors, reviewers, and colleagues at Google AI, #MPI-INF and #MPI-SWS for their valuable feedback!

We have updated our recent post on measuring gendered correlations in pre-trained #NLP models to include Zari, a series of checkpoints to identify and reduce gendered correlations while maintaining state-of-the-art accuracy on standard NLP metrics. github.com/google-researc…

Honored to win the Test of Time Award (cikmconference.org/cikmToTA.html) in CIKM2020, w/ my Ph.D. advisor NaN and my buddy Kyumin Lee. An amazing time working with these two gents, and never saw it coming from the first ever paper (people.engr.tamu.edu/caverlee/pubs/…) in grad school.

NEW from a big collaboration at Google: Underspecification Presents Challenges for Credibility in Modern Machine Learning Explores a common failure mode when applying ML to real-world problems. 🧵 1/14 arxiv.org/abs/2011.03395

Announcing the release of MinDiff, a new regularization technique available in the TensorFlow Model Remediation library for effectively and efficiently mitigating unfair biases when training #MachineLearning models. Learn more below: goo.gle/3f3Qacu

As we develop #ResponsibleAI techniques, we test them out in our products, and then work to share our learnings Excited to launch the TF Model Remediation Library, based on some amazing research with Alex Beutel Jilin Chen Ed H. Chi and more ai.googleblog.com/2020/11/mitiga…

A new significant release from my team at Google for inclusive and responsible ML. In practice, the MinDiff technique has been great for improving inclusiveness in our ML models, and the use of MMD (maximum mean discrepancy) was important to making it easy to apply.

Could not agree more! Tina Eliassi has been a key mentor throughout my career! Evimaria's early advice on proposal writing has been instrumental. Also Leman Akoglu with whom I co-authored one of my favorite papers while at CMU, and *of course* Tanya Berger-Wolf (not on twitter)