HiMed releases a Hindi medical reasoning corpus and benchmark and shows that training an 8B LLM with decaying scaffolding reward improves Hindi performance and narrows the English-Hindi accuracy gap.
arXiv preprint arXiv:2407.19655
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
HiMed: Incentivizing Hindi Reasoning in Medical LLMs
HiMed releases a Hindi medical reasoning corpus and benchmark and shows that training an 8B LLM with decaying scaffolding reward improves Hindi performance and narrows the English-Hindi accuracy gap.