Research
Work in multilingual and multimodal intelligence, foundation model architectures, reasoning and agentic systems, and evaluation and trustworthy AI.
Theme
Language diversity, code-mixing, and multimodal understanding.
Projects
Hindi–English benchmarks and tools, 2023–2026.
Page → Funded project · ANRFMultimodal datasets and foundation models, 2026–2029.
Page →Resources and papers
Theme
New architectures, efficient computation, and model adaptation.
Projects
India-centric foundation LLM programme.
Page → ProjectGanga: open Hindi language models.
Page →Resources and papers
Theme
Reasoning, procedural execution, world models, and grounded intelligence.
Resources and papers
Theme
Evaluation, alignment, robustness, security, and interpretability.
Projects
Eka-Eval: Indian-language LLM evaluation suite.
Page → ProjectUnityAI-Guard: multilingual toxicity detection for Indian languages.
Page →Resources and papers
Code-generation evaluation.
Page → BenchmarkScientific leaderboard generation.
Page → DatasetPersona–instruction alignment.
Page → PaperCross-lingual backdoor attacks.
Page → PaperCross-lingual model editing.
Page → DatasetMultimodal cultural toxicity.
Page →
Past
NLPExplorer, Gandhipedia, and scholarly document AI.
Projects
Portal for NLP papers, authors, and venues.
Page → Funded project · NCSMPortal for Gandhian literature and networks.
Page → Funded project · SERBTable and chart extraction, 2019–2022.
Page →Resources and papers
Scholarly PDF extraction.
Page → DatasetPeer-review comparisons.
Page → PaperScientific-article embeddings.
Page → DatasetTextbook question–answer pairs.
Page → PaperPopularity vs performance.
Page → PaperAnnotator disagreement.
Page → Dataset2019 election WhatsApp corpus.
Page →