DSBC// Blog
Writing on multilingual evaluation, low-resource language systems, agent benchmarking, and the practical side of research work.
// Featured EssayI did not start thinking about research from Kathmandu as an identity statement. It entered my life as a practical condition. It was the place from which I was reading, missing deadlines, catching up, trying again, taking buses, saving papers offline, and learning how much distance can sharpen your eye instead of only slowing you down.
A research note on our CHiPSAL paper about Devanagari language identification, hate speech detection, and hate-target detection, written around one core argument: shared script does not turn those tasks into one coherent modeling problem.