Building a Safety Benchmark Dataset in Hindi
This project has concluded.
In 2024, Tattle was selected in this pilot project to build a dataset of prompts in Hindi as part of ML Commons' safety benchmark. We followed Uli's participatory approach and created 2000 prompts in Hindi on two hazard categories: hate and sex-related crimes. These prompts were created by an expert group, consisting of individuals with expertise in journalism, social work, feminist advocacy, gender studies, fact-checking, political campaigning, education, psychology, and research.
Project Team
Mansi Gupta
Srravya C
Vamsi Krishna Pothuru
Saumya Gupta
Tarunima Prabhakar
Aatman Vaidya
Kaustubha Kalidindi
Denny George
Maanas B
Outcomes
AI Safety Benchmark Datasets in Hindi
Analysis of Indic Language Capabilities in LLMs