Webinar | Gender Biases in LLMs: Perspectives from the Indian Context
Between 2023-24, with support from the Gates Foundation, Digital Futures Lab conducted extensive research to investigate the sources of gender bias in large language models (LLMs) built and customised for social sector use cases in India.
As part of this effort, we have also developed a set of design principles and indicative strategies for de-biasing these models, specifically tailored to the realities of the Indian context.
This webinar formally launched learnings, perspectives, and strategies from the various research outputs that have emerged from our year-long work. These include a comprehensive guidebook on mitigating gender biases in Indian Language LLMs, a primer on the burgeoning LLM ecosystem in India, and a guide to gender-sensitive user research, among other outputs.
The webinar also had a keynote address from Mr Abhishek Singh, Additional Secretary, Ministry of Electronics and Information Technology, Govt of India. Digital Futures Lab's Founder and Executive Director, Urvashi Aneja led and moderated a panel discussion after the presentation of findings by Aarushi Gupta.
The panel featured experts from across the Generative AI ecosystem in India:
- Anaita Singh of Gates Foundation India
- Danish Pruthi of IISc, Bangalore
- Kalika Bali of Microsoft Research India
- Nidhi Bhasin of Digital Green Trust
- Safiya Husain of Karya
- Saurabh Karn of Sarvam AI