Independent researcher working on language model evaluation, multilingual reasoning, and reproducible AI. I am researching on how large language models reason across languages — especially Indic and low-resource languages.
Current work: measuring the cross-lingual reasoning gap — how much LLM reasoning accuracy changes when problems are posed in Hindi, Sanskrit, and Tamil versus English, and whether chain-of-thought prompting helps equally across languages.
Interests: LLM evaluation · multilingual NLP · reasoning · low-resource languages · reproducibility
Find me
- ORCID: https://orcid.org/0009-0002-7453-2897
- Kaggle: https://kaggle.com/researchai2026
- Hugging Face: https://huggingface.co/stresearch-dev
Open to reproducible, open-source research collaborations.