CORDIS Project
Funding breakdown and partner intelligence are Premium
Sign in and upgrade to Premium for EU contribution totals, consortium analytics, OpenAlex research context, and AI summaries. · 0 consortium intelligence fields visible of 1
Start free • Cancel anytime • 14-day refund guarantee
The DISHARM project aims to create a framework for evaluating data filtering strategies to reduce harmful content in training datasets for large language models. It emphasizes collaboration with activists to assess the impact on vulnerable identities and promote fairness in technology.
Large Language Models (LLMs) learn everything they know of the World from what they find in training datasets: if datasets include harmful content, it is more likely that they learn how to produce discriminating outputs.
Therefore, reducing the presence of harmful contents in the input training dataset is a crucial step to develop safer and fairer technologies.
However, the effectiveness of existing data filtering strategies for harm reduction is still an understudied topic in NLP research. DISH…
IT-UNIVERSITETET I KOBENHAVN
Partner organizations (coordinator is shown above), with normalized type and CORDIS activity type. Guests see up to 4 partners.
Similar projects, consortium collaboration history, frequent partners, and OpenAlex research context.
Guests see up to 5 EuroSciVoc fields.
Guests see up to 5 topics.
Guests see up to 5 keywords.
