Mounika Marredy

I am currently a postdoctoral researcher at university of Bonn under the supervision of Prof. Lucie Flek. My research interests lies in interpretation, explainability, and robustness of large language models. I also focus on modelling dynamic user representations and user opinion beliefs. Additionally, I am interested in understanding the working principals of large language models by aligning them with brain datasets (fMRI, MEG). In my PhD, I worked on creating resources and pretrained language models for low-resource Indian languages. In particular, I worked on creating different annotated datasets for tasks like sentiment analysis, emotion identification, hate speech, sarcasm, clickbait detection, and NER where I collected the data from online media. I have publications at top conference venues like ACL, NAACL, EMNLP, INTERSPEECH, ICASSP, Cogsci, and IJCNN.

Areas of Interest

  - Robustness and Interpretability of Large Language and Multi-lingual models 
  - Modelling dynamic user representations
  - Low resource NLP
  - Language analysis in the brain