Tag
Terminal-Bench-LILT presents a multilingual coding benchmark with 300 authentic tasks in 10 languages, exposing gaps in AI models' handling of non-English issues like internationalization and cultural conventions.
The paper investigates whether large language models can phonetically decode encoded languages, such as German written in Cyrillic characters, to assess their abstraction capabilities and performance beyond standard Latin-based training data.
The article explores the hypothesis that AI writing tools are subtly flattening individual writing voices into a shared average style, potentially reducing stylistic diversity over time.
This paper introduces a mathematical framework to study how reliance on shared LLMs for writing may reduce population-level linguistic diversity, analyzing fixed, recursive, and personalized interaction mechanisms and characterizing equilibria and convergence rates.
Nonprofit Current AI is building open, public AI infrastructure to serve underserved languages and communities, backed by $400M in funding from governments and foundations. It has launched Suno Sutra, an offline AI device supporting 22 Indian languages, and is funding open-source AI projects.