Adit Radhakrishanan is a first year PhD student in EECS at MIT, advised by Professor Caroline Uhler. He received his BS and MEng from MIT. Adit is interested in deep learning theory, interpretable neural models, and causal inference.
His current research focuses on understanding how memorization (as opposed to interpolation) occurs in deep learning. Further, he is interested in how these results impact generalization and interpretability of the learned models.