Machine-and-Human Intelligence Lab

We are a research group at BITS Pilani, Goa Campus working on AI for Science. Our research focuses on bridging the gap between human expertise and machine learning (ML). Rather than relying solely on data-driven learning, we integrate scientific domain knowledge directly into ML models, including deep neural networks. This improves both predictive performance and explainability, enabling AI systems that are not only accurate but also understandable, trustworthy, and useful to scientists.

Our primary focus areas are these:

Neurosymbolic AI: integrating deep neural networks with symbolic machine learning techniques such as Inductive Logic Programming to build models that are accurate, interpretable, and data-efficient; neurosymbolic inclusion of domain-knowledge into deep networks.
Deep learning: graph neural networks, sparse neural networks, sequence models, transformers, LLMs, uncertainty quantification, and explainability, with a focus on reliability for scientific applications.
AI for Science: AI for gene regulation, genomic foundation models, multi-omics cancer analysis, biomarker discovery, and early-stage drug discovery.
Causal ML: exploring causal machine learning methods that go beyond correlation to discover causal structure in biological and clinical data.

News

05 Jun 2026: pVR paper is out. Check the preprint on arXiv.
28 May 2026: BIRDNet paper is out. Check the preprint on arXiv.
Apr 2026: Join us, if you want to work on AI for Science. See opportunities