John Hoffman

TL;DR: A family of models that enable end-to-end expressive and multilingual translations in a streaming fashion, including the first known red-teaming effort for multimodal machine translation, a system for the detection and mitigation of added toxicity, a systematic evaluation of gender bias, and an inaudible localized watermarking mechanism designed to dampen the impact of deepfakes are introduced.

TL;DR: The first multilingual system capable of translating from and into English for both speech and text, SeamlessM4T is developed, which sets a new standard for translations into multiple target languages and is evaluated on gender bias and added toxicity to assess translation safety.

TL;DR: In both offline ad bidding simulations and testing on live traffic, the methods proved to be effective in simultaneously controlling multiple KPIs, and bringing them toward their respective goals.

TL;DR: In this paper, the authors present Automatic Model Selector (AMS), a system for scalable online selection of RTB bidding strategies based on real-world performance metrics, which employs Multi-Armed Bandits (MAB) to near-simultaneously run and evaluate multiple models against live traffic, allocating the most traffic to the best-performing models while decreasing traffic to those with poorer online performance.