Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

Picture for Tal Linzen

The Language Model Understood the Prompt was Ambiguous: Probing Syntactic Uncertainty Through Generation

Sep 16, 2021
Laura Aina, Tal Linzen

* To appear in Proceedings of BlackboxNLP 2021: Analyzing and Interpreting Neural Networks for NLP 

  Access Paper or Ask Questions

Frequency Effects on Syntactic Rule Learning in Transformers

Sep 14, 2021
Jason Wei, Dan Garrette, Tal Linzen, Ellie Pavlick

* Camera ready for EMNLP 2021 

  Access Paper or Ask Questions

NOPE: A Corpus of Naturally-Occurring Presuppositions in English

Sep 14, 2021
Alicia Parrish, Sebastian Schuster, Alex Warstadt, Omar Agha, Soo-Hwan Lee, Zhuoye Zhao, Samuel R. Bowman, Tal Linzen

* CoNLL 2021. Data and code available at 

  Access Paper or Ask Questions

The MultiBERTs: BERT Reproductions for Robustness Analysis

Jun 30, 2021
Thibault Sellam, Steve Yadlowsky, Jason Wei, Naomi Saphra, Alexander D'Amour, Tal Linzen, Jasmijn Bastings, Iulia Turc, Jacob Eisenstein, Dipanjan Das, Ian Tenney, Ellie Pavlick

* Checkpoints and example analyses: 

  Access Paper or Ask Questions

Causal Analysis of Syntactic Agreement Mechanisms in Neural Language Models

Jun 22, 2021
Matthew Finlayson, Aaron Mueller, Sebastian Gehrmann, Stuart Shieber, Tal Linzen, Yonatan Belinkov

* Accepted to ACL-IJCNLP 2021 

  Access Paper or Ask Questions

Counterfactual Interventions Reveal the Causal Effect of Relative Clause Representations on Agreement Prediction

May 19, 2021
Shauli Ravfogel, Grusha Prasad, Tal Linzen, Yoav Goldberg

* Equal contribution by SR and GP. Uploaded new version to fix typo in the abstract 

  Access Paper or Ask Questions

Evaluating Groundedness in Dialogue Systems: The BEGIN Benchmark

Apr 30, 2021
Nouha Dziri, Hannah Rashkin, Tal Linzen, David Reitter

* 12 pages, 2 figures 

  Access Paper or Ask Questions

Does Putting a Linguist in the Loop Improve NLU Data Collection?

Apr 15, 2021
Alicia Parrish, William Huang, Omar Agha, Soo-Hwan Lee, Nikita Nangia, Alex Warstadt, Karmanya Aggarwal, Emily Allaway, Tal Linzen, Samuel R. Bowman

* 14 pages, 10 figures 

  Access Paper or Ask Questions

COGS: A Compositional Generalization Challenge Based on Semantic Interpretation

Oct 12, 2020
Najoung Kim, Tal Linzen

* Accepted to EMNLP 2020 

  Access Paper or Ask Questions

Universal linguistic inductive biases via meta-learning

Jun 29, 2020
R. Thomas McCoy, Erin Grant, Paul Smolensky, Thomas L. Griffiths, Tal Linzen

* To appear in the Proceedings of the 42nd Annual Conference of the Cognitive Science Society 

  Access Paper or Ask Questions

Cross-Linguistic Syntactic Evaluation of Word Prediction Models

May 21, 2020
Aaron Mueller, Garrett Nicolai, Panayiota Petrou-Zeniou, Natalia Talmina, Tal Linzen

* Accepted for presentation at ACL 2020 

  Access Paper or Ask Questions

How Can We Accelerate Progress Towards Human-like Linguistic Generalization?

May 03, 2020
Tal Linzen

* ACL 2020 

  Access Paper or Ask Questions

Representations of Syntax [MASK] Useful: Effects of Constituency and Dependency Structure in Recursive LSTMs

Apr 30, 2020
Michael A. Lepori, Tal Linzen, R. Thomas McCoy

* To appear in Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (ACL-2020) 

  Access Paper or Ask Questions

Syntactic Data Augmentation Increases Robustness to Inference Heuristics

Apr 24, 2020
Junghyun Min, R. Thomas McCoy, Dipanjan Das, Emily Pitler, Tal Linzen

* ACL 2020 

  Access Paper or Ask Questions

Syntactic Structure from Deep Learning

Apr 22, 2020
Tal Linzen, Marco Baroni

* In press at Annual Reviews of Linguistics 

  Access Paper or Ask Questions

Does syntax need to grow on trees? Sources of hierarchical inductive bias in sequence-to-sequence networks

Jan 10, 2020
R. Thomas McCoy, Robert Frank, Tal Linzen

* 12 pages, 10 figures; accepted to TACL 

  Access Paper or Ask Questions

Discovering the Compositional Structure of Vector Representations with Role Learning Networks

Nov 17, 2019
Paul Soulos, Tom McCoy, Tal Linzen, Paul Smolensky

  Access Paper or Ask Questions

BERTs of a feather do not generalize together: Large variability in generalization across models with similar test set performance

Nov 07, 2019
R. Thomas McCoy, Junghyun Min, Tal Linzen

* 7 pages, 4 figures 

  Access Paper or Ask Questions

Using Priming to Uncover the Organization of Syntactic Representations in Neural Language Models

Sep 23, 2019
Grusha Prasad, Marten van Schijndel, Tal Linzen

* CoNLL 2019 
* 9 pages paper, 2 pages references and 3 pages supplementary materials. Code for the templates and analyses can be found here: 

  Access Paper or Ask Questions

Quantity doesn't buy quality syntax with neural language models

Aug 31, 2019
Marten van Schijndel, Aaron Mueller, Tal Linzen

* Accepted for presentation at EMNLP-IJCNLP 2019 

  Access Paper or Ask Questions

Probing What Different NLP Tasks Teach Machines about Function Word Comprehension

Apr 25, 2019
Najoung Kim, Roma Patel, Adam Poliak, Alex Wang, Patrick Xia, R. Thomas McCoy, Ian Tenney, Alexis Ross, Tal Linzen, Benjamin Van Durme, Samuel R. Bowman, Ellie Pavlick

* Accepted to *SEM 2019 

  Access Paper or Ask Questions

Analyzing and Interpreting Neural Networks for NLP: A Report on the First BlackboxNLP Workshop

Apr 05, 2019
Afra Alishahi, Grzegorz Chrupała, Tal Linzen

  Access Paper or Ask Questions

Studying the Inductive Biases of RNNs with Synthetic Variations of Natural Languages

Mar 26, 2019
Shauli Ravfogel, Yoav Goldberg, Tal Linzen

* Accepted as a long paper in NAACL 2019 

  Access Paper or Ask Questions

Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference

Feb 04, 2019
R. Thomas McCoy, Ellie Pavlick, Tal Linzen

* 8 pages plus references and appendices 

  Access Paper or Ask Questions

Human few-shot learning of compositional instructions

Jan 14, 2019
Brenden M. Lake, Tal Linzen, Marco Baroni

  Access Paper or Ask Questions

RNNs Implicitly Implement Tensor Product Representations

Dec 20, 2018
R. Thomas McCoy, Tal Linzen, Ewan Dunbar, Paul Smolensky

* 8 pages plus references and appendices 

  Access Paper or Ask Questions