Research
Projects I'm currently working on, and ones I've wrapped up.
Current Projects
Dependency Distance in Code-Mixing
Studying dependency distance in code-mixed text, and what it reveals about the syntactic processing cost of switching languages mid-sentence.
A Multilingual fMRI/EEG Dataset for Indian English, Hindi, and Telugu
Building a neuroimaging dataset spanning Indian English, Hindi, and Telugu, for studying language processing in the brain.
Reasoning Failures in Vision-Language Models
Investigating where and why vision-language models break down on multi-step reasoning tasks.
Past Projects
Adapting Multilingual Models to Code-Mixed Tasks via Model Merging
Investigating effective utilization of unlabeled code-mixed datasets and monolingual resources via model merging.
- Collaboration with Max Planck Institute, MBZUAI, and IIIT Hyderabad
- Accepted at CODS-IKDD 2025
SceneGraMMi: Scene Graph-boosted Hybrid-fusion for Multi-Modal Misinformation
A scene-graph-boosted hybrid-fusion approach for multi-modal misinformation veracity prediction.
arXiv:2410.15517