Hi, I’m Debangan 🎓

I’m an undergraduate researcher working on how to make AI systems and models safer for all. My interests lie in interpretability, explainability, and machine unlearning. I am also fascinated by multimodal models, NLP and geometric deep learning. I enjoy building principled approaches to make AI systems more transparent, secure, and aligned with human intent. Cheers to a safer and more responsible future of artificial intelligence 🥂.

🔗 Links

📚 Publications

• A Cognac shot to forget bad memories: Corrective Unlearning in GNNs
  V. Kolipaka, A. Sinha, D. Mishra, S. Kumar, A. Arun, S. Goel, P. Kumaraguru, International Conference on Machine Learning (ICML), 2025. [PDF]
• SANITY CHECKS FOR EVALUATING GRAPH UNLEARNING
  V. Kolipaka, A. Sinha, D. Mishra, S. Kumar, A. Arun, S. Goel, P. Kumaraguru, CoLLAs Workshop Track, 2024. [PDF]