Séminaires RALI-OLST

Searching for Needles in a Haystack: On the Role of Incidental Bilingualism in PaLM’s Translation Capability

George Foster (geord (point) foster <at> gmail (point) com)

Google Montréal

Wednesday 6 September 2023 at 11:30 AM

Salle 3195, Pavillon André-Aisenstadt — Diffusion simultanée sur Zoom


Large, multilingual language models exhibit surprisingly good zero- or few-shot machine translation capabilities, despite having never seen the intentionally-included translation examples provided to typical neural translation systems. We investigate the role of incidental bilingualism—the unintentional consumption of bilingual signals, including translation examples—in explaining the translation capabilities of large language models, taking the Pathways Language Model (PaLM) as a case study. We introduce a mixed-method approach to measure and understand incidental bilingualism at scale. We show that PaLM is exposed to over 30 million translation pairs across at least 44 languages. Furthermore, the amount of incidental bilingual content is highly correlated with the amount of monolingual in-language content for non-English languages. We relate incidental bilingual content to zero-shot prompts and show that it can be used to mine new prompts to improve PaLM’s out-of-English zero-shot translation quality. Finally, in a series of small-scale ablations, we show that its presence has a substantial impact on translation capabilities, although this impact diminishes with model scale.

Recording: https://drive.google.com/file/d/1IktooXw8vs9IZtW5CzOMYhphHL27jzlj/view?usp=drive_link



Join with Zoom at this url.
Meeting ID: 916 9097 5818, 𝑷𝒂𝒔𝒔𝒄𝒐𝒅𝒆: 343273.
Phone numbers: https://umontreal.zoom.us/u/abitNZzLg.
One tap mobile: +14388097799,,91690975818#,,,,,,0#,,343273#


Follow this link to subscribe to future RALI-OLST announcements.
http://rali.iro.umontreal.ca/rali/?q=fr/node/1631

See all the weekly talks for the year:

1991 1992 1993 1994 1995 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024