Skip to main content

The CRITICS project on “Critical Science Without Borders: LLMs for Translation of Scientific Knowledge in Multilingual Contexts” proposes to transform the future of science accessibility and literacy through the convergence of advanced Machine Translation based on Large Language Models (LLMs) and educational technology. By developing specialized LLMs for scientific content translation, the project aims to make cutting-edge scientific knowledge accessible in local languages while maintaining technical accuracy and disciplinary rigor. The project integrates three interconnected technological innovations: advanced Machine Translation systems optimized for scientific documents, AI-powered generation of Teaching Learning Sequences (TLSs) that incorporate local contexts while maintaining scientific standards, and automated assessment tools for evaluating critical thinking and scientific reasoning in students' native languages. CRITICS tackles significant technical challenges including domain-specific terminology machine translation, analysis of scientific argumentation across languages, and automated feedback in competency-based assessment through the generation of critical questions. The project's impact extends beyond technology development through a robust dissemination and exploitation strategy. This includes open-source release of LLMs, tools, and resources; collaboration with educational institutions for testing of the implementations; engagement with policymakers; and public outreach activities. Through these efforts, CRITICS aims to democratize access to scientific knowledge, enhance critical thinking, and reduce linguistic inequities in scientific literacy. By enabling students worldwide to engage with scientific concepts in their own languages, CRITICS represents a significant step toward a more inclusive global scientific community where language is no longer a barrier to participation in scientific discourse and advancement.

Call Topic: Science in your Own Language (SOL), Call 2025
Start date: (36 months)
Funding support: 936 117 €