Tarun Tater

I'm a final-year Ph.D. researcher in the MUDCAT Project supervised by Prof. Dr. Sabine Schulte im Walde (Uni Stuttgart) and Dr. Diego Frassinelli (LMU), working on multimodal representations of abstract and concrete concepts. My research explores how abstract and concrete concepts differ, with a focus on how they are represented and aligned across language and vision. I'm especially interested in how computational models — such as Vision-Language Models (VLMs) like CLIP and SigLIP — process semantic information, and how their representations compare to human cognition.

Prior to my Ph.D., I worked as a Senior Staff Research Engineer at IBM Research, India, contributing to both research and product-oriented projects. These included tasks such as Text-to-SQL conversion, sarcasm generation, and automation of invoice processing systems. I hold an integrated B.Tech + M.Tech degree in Information Technology from IIIT Bangalore, where I was awarded the gold medal for all-round excellence.

My current work involves analyzing computational models that are better aligned with how humans represent and interpret abstract and concrete concepts.

Research Interests  · Concreteness & Abstractness · Multimodal NLP · Vision-Language Models · Text Generation 

Project Page:  Multimodal Dimensions and Computational Applications of Abstractness (MUDCAT)
Group Page: Semrel Website