Our paper: "Gromov-Wasserstein Alignment of Word Embedding Spaces" is now available (arxiv.org/abs/1809.00013). TL;DR: The Gromov-Wasserstein distance provides a simple, principled objective to align (w/o supervision) word embedding spaces, even of different dimensionality!

Congratulations Tal! old account

Our paper "GraphIE: A Graph-Based Framework for Information Extraction" has been accepted to #NAACL2019. We study how to model the graph structure of the data in various IE tasks. Joint work with @esantus Jiang Guo Zhijing Jin and Regina Barzilay. (arxiv.org/abs/1810.13083)

The new shared task in QA at @MRQA2019 Workshop is released

Development datasets released! 6 in-domain and 6 out-of-domain including BioASQ, DROP, DuoRC, RACE, RelationExtraction, TextbookQA! Also released BERT baseline results. All the information at github.com/mrqa/MRQA-Shar…. Check out and let us know if you have questions! #mrqa2019

Our #emnlp2019 paper is now on arxiv:arxiv.org/abs/1908.05267 * Extending #FEVER (fact-checking) eval dataset to eliminate bias. * Regularizing the training to alleviate the bias. Coauthors: Darsh Shah, Serene Yeo, Daniel Filizzola, @esantus, Regina Barzilay @emnlp2019 #nlproc

Few-shot Text Classification with Distributional Signatures. What happens if you take meta-learning for vision and apply it to NLP? Prototypical Networks with lexical features perform worse than nearest neighbors on new classes. How can we do better? ;) arxiv.org/abs/1908.06039

Are we protected from GPT-2, #GROVER style models generating fake content? What happens if they are also used legitimately as writings assistants? Check our new report: arxiv.org/abs/1908.09805 with Roei Schuster, @Darsh71307636, Regina Barzilay. #NLProc #emnlp2019 #FakeNews #GPT2

Check-out our new paper - arxiv.org/pdf/1909.13838… Automatic Fact-guided sentence modification. Method to automatically modify the factual information in a sentence. Joint work with old account , Prof. Regina Barzilay.

If you're at @emnlp2019, don't miss our talks: Towards Debiasing Fact Verification Models * Wednesday 15:42 (2B) * Tal Schuster Darsh J Shah Working Hard or Hardly Working: Challenges of Integrating Typology into Neural Dependency Parsers * Thursday 15:30 (201A) * Adam Fisch

#NeurIPS2019 Our work with MIT improves the interpretability of NLP models with an adversarial class-wise rationalization technique, which can find explanations towards any given class. Poster: Tue @ East Exhibition Hall B + C #1. MIT-IBM Watson AI Lab David Cox MIT CSAIL Mo Yu

Congratulations!

In our IEEE S&P paper, led by Roei Schuster, we control the embeddings of words by introducing minimal changes to the pretraining data (e.g. #Wiki edits). This #word_embeddings attack affects many downstream #NLProc tasks! Cornell Tech MIT CSAIL link: arxiv.org/abs/2001.04935

Is your Fact Verification model robust enough? Consider adding #VitaminC 🍊 Check out our new #NAACL2021 paper: "Get Your Vitamin C! Robust Fact Verification with Contrastive Evidence" with Adam Fisch and Regina Barzilay 🔗 arxiv.org/abs/2103.08541 #NLProc #FakeNews📰 🧵1/N

New #NAACL2021 paper out on robust fact verification. Sources like Wikipedia are continuously edited with the latest information. In order to keep up, our models need to be sensitive to these changes in evidence when verifying claims. Work with Tal Schuster and Regina Barzilay!

New Preprint with Adam Fisch, T.Jaakkola and Regina Barzilay. We present Consistent Accelerated Inference via 𝐂onfident 𝐀daptive 𝐓ransformers (CATs) CATs can speed up inference 😺 while guaranteeing consistency 😼. The code is available🙀 🔗people.csail.mit.edu/tals/static/Co… #NLProc

Large pre-trained Transformers are great, but expensive to run. But making them more efficient (e.g., early exits) can give undesirable performance hits. In our new work, we speed up inference while guaranteeing consistency with the original model up to a specifiable tolerance.

Need to debias your new task? Learn how, from your old one. Check out our #ICML2022 paper "Learning Stable Classifiers by Transferring Unstable Features" with Shiyu Chang and Regina Barzilay Paper -> arxiv.org/abs/2106.07847 Code -> github.com/YujiaBao/tofu

I’m thrilled to announce Conformal Risk Control: a way to bound quantities other than coverage with conformal prediction. arxiv.org/abs/2208.02814 Check out the worked examples in CV and NLP! The best part is: it’s exactly the same algorithm as split conformal prediction🤯🧵1/5