S
Suchin Gururangan
Researcher at University of Washington
Publications - 30
Citations - 3807
Suchin Gururangan is an academic researcher from University of Washington. The author has contributed to research in topics: Computer science & Hyperparameter. The author has an hindex of 10, co-authored 21 publications receiving 2021 citations. Previous affiliations of Suchin Gururangan include Carnegie Mellon University & Allen Institute for Artificial Intelligence.
Papers
More filters
Proceedings ArticleDOI
Don't Stop Pretraining: Adapt Language Models to Domains and Tasks
Suchin Gururangan,Ana Marasović,Ana Marasović,Swabha Swayamdipta,Kyle Lo,Iz Beltagy,Doug Downey,Noah A. Smith,Noah A. Smith +8 more
TL;DR: It is consistently found that multi-phase adaptive pretraining offers large gains in task performance, and it is shown that adapting to a task corpus augmented using simple data selection strategies is an effective alternative, especially when resources for domain-adaptive pretraining might be unavailable.
Proceedings ArticleDOI
Annotation Artifacts in Natural Language Inference Data
Suchin Gururangan,Swabha Swayamdipta,Omer Levy,Roy Schwartz,Roy Schwartz,Samuel R. Bowman,Noah A. Smith +6 more
TL;DR: The authors showed that a simple text categorization model can correctly classify the hypothesis alone in about 67% of SNLI and 53% of MultiNLI, showing that specific linguistic phenomena such as negation and vagueness are highly correlated with certain inference classes.
Posted Content
Annotation Artifacts in Natural Language Inference Data
Suchin Gururangan,Swabha Swayamdipta,Omer Levy,Roy Schwartz,Roy Schwartz,Samuel R. Bowman,Noah A. Smith +6 more
TL;DR: It is shown that a simple text categorization model can correctly classify the hypothesis alone in about 67% of SNLI and 53% of MultiNLI, and that specific linguistic phenomena such as negation and vagueness are highly correlated with certain inference classes.
Proceedings ArticleDOI
RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models
TL;DR: It is found that pretrained LMs can degenerate into toxic text even from seemingly innocuous prompts, and empirically assess several controllable generation methods find that while data- or compute-intensive methods are more effective at steering away from toxicity than simpler solutions, no current method is failsafe against neural toxic degeneration.
Proceedings ArticleDOI
Show Your Work: Improved Reporting of Experimental Results
TL;DR: It is demonstrated that test-set performance scores alone are insufficient for drawing accurate conclusions about which model performs best, and a novel technique is presented: expected validation performance of the best-found model as a function of computation budget.