London

Applying machine learning (ML) to electronic health records (EHRs) has shown promise in predicting clinical outcomes. However, many of these ML models are based on single-task learning instead of multi-task models, which take into account a variety of competing risks along with the interdependencies between organ systems that factor into patient outcomes in a realistic setting. In “Multi-task prediction of organ dysfunction in the ICU using sequential sub-network routing”, we propose a multi-task learning (MTL) architecture, called Sequential Sub-Network Routing (SeqSNR), that better captures the complexity of a realistic setting.

Composing longer pieces of text on smartphones is still quite cumbersome - grammatical errors can often creep into the text and correcting them can be time consuming on a small display with limited controls. To address this, we launched a grammar correction feature that is directly built into Gboard on Pixel 6 that works entirely on-device to preserve privacy, detecting and suggesting corrections for grammatical errors while the user is typing.

The conversations we might observe with today’s digital assistants don't reach the level of dialog complexity we need to model human-level understanding. To address this, we released the Coached Conversational Preference Elicitation (CCPE) and Taskmaster-1 English dialog datasets.

In “Universal Transformers” we extend the standard Transformer to be computationally universal (Turing complete) using a novel, efficient flavor of parallel-in-time recurrence which yields stronger results across a wider range of tasks.

Making synthesized speech sound as natural as possible is difficult, partly because evaluation is normally performed on a sentence-by-sentence basis. This new research compares several ways of evaluating synthesized speech for multi-line texts.

Due to the COVID-19 pandemic, scientists around the world are publishing an immense amount of new research in order to understand and combat the disease. While the volume of research is very encouraging, it can be difficult to keep up with the rapid pace of new publications. To help address this problem, we launched the COVID-19 Research Explorer, a tool that uses Natural Language Understanding algorithms that helps scientists and researchers efficiently pore through articles for answers or evidence to COVID-19-related questions.