LaMDA: Towards Safe, Grounded, and High-Quality Dialog Models for Everything
Evaluating Syntactic Abilities of Language Models
Grammar Correction as You Type, on Pixel 6
Two New Datasets for Conversational NLP: TimeDial and Disfl-QA
Constructing Transformers For Longer Sequences with Sparse Attention Methods
RxR: A Multilingual Benchmark for Navigation Instruction Following
Navigating Recorder Transcripts Easily, with Smart Scrolling
Measuring Gendered Correlations in Pre-trained NLP Models
Language-Agnostic BERT Sentence Embedding
SmartReply for YouTube Creators
Evaluating Natural Language Generation with BLEURT
