40 matched 310M decoder-only models from the Double Trouble study of bilingual pretraining. Paper: arxiv.org/abs/2608.26576
Anjishnu Mukherjee
iamshnoo
AI & ML interests
Natural Language Processing, Multilingual NLP, Large Language Models
Recent Activity
updated a dataset 1 day ago
iamshnoo/LocalNewsQA authored a paper 2 days ago
Purdah and Patriarchy: Evaluating and Mitigating South Asian Biases in Open-Ended Multilingual LLM Generations authored a paper 2 days ago
BiasDora: Exploring Hidden Biased Associations in Vision-Language Models