Software Engineer • NLP Research Enthusiast
I'm a Software Engineer with research interests in Natural Language Processing (NLP) and Machine Learning. My work lies at the intersection of software engineering and modern NLP, where I explore how language models can be made more reliable, secure, and trustworthy. I'm particularly interested in developing robust LLM-based systems through representation learning, adversarial robustness, retrieval-augmented generation, and privacy-preserving machine learning.
- Natural Language Processing (NLP)
- Trustworthy Large Language Models
- Adversarial Robustness of Language Models
- Retrieval-Augmented Generation (RAG)
- Representation Learning
- Hallucination & Factuality
- Multilingual NLP
- Low-Resource NLP
- Privacy-Preserving Machine Learning (PPML)
Tensor Decomposition for Adversarial Prompt Injection Detection in Retrieval-Augmented Generation Systems
Investigating tensor decomposition-based representation learning methods to improve the robustness of LLM-powered applications against adversarial attacks.
📖 Medium: https://medium.com/@subisurbiee
Build, Learn, Iterate.