Rishi Singhal

North Carolina State University, Raleigh

prof_pic.jpg

Engineering Building II (EB2), Room 3224

Box 8206 NCSU Campus

Raleigh, NC 27695

Hi, my name is Rishi Singhal. I am a first-year Ph.D. student in Computer Science at North Carolina State University Department of Computer Science, Raleigh. I am working under the guidance of Dr. Jung-Eun Kim. I completed my M.S. in Computer Science at NC State in Spring 2025. Prior to that, I graduated with a B.Tech. in Electronics & Communications Engineering (ECE) from Indraprastha Institute of Information Technology (IIIT), Delhi in 2023. During my undergraduate studies, I had the opportunity to collaborate with Dr. Rajiv Ratn Shah.

My research is driven by a deep interest in the dynamics of memorization and generalization in deep learning systems. Specifically, I study how models encode these phenomena, why certain features tend to be memorized while others generalize, and how to design architectures or training methods that promote generalization while mitigating memorization.

Beyond this core focus, my broader research interests include explainability and interpretability in AI, addressing spurious correlations and invariances, and developing approaches that enhance the efficiency, robustness, safety, and privacy of large-scale foundation models. My long-term goal is to contribute toward building trustworthy AI systems that balance strong generalization performance with privacy safety and computational efficiency.

news

Sep 18, 2025 Paper titled “Impact of LayerNorm on Memorization and Learning in Transformers” accepted at **NeurIPS 2025** Main Conference! 🚀✨
Aug 15, 2025 Started PhD in Computer Science at NC State University, Raleigh, under the guidance of Dr. Jung-Eun Kim :sparkles: :smile:
May 15, 2025 Completed Masters in Computer Science from NC State University, Raleigh under the supervision of Dr. Jung-Eun Kim.
Jun 15, 2024 Worked at Fermilab as a Machine Learning Intern as part of the MicroBoone Project from May 2024 - August 2024.

selected publications

  1. Impact of Layer Normalization on Memorization and Learning in Transformers
    Rishi Singhal and Jung-Eun Kim
    Advances in Neural Information Processing Systems, 2025
  2. iclr.png
    Beyond words: A topological exploration of coherence in text documents
    Samyak Jain, Rishi Singhal, Sriram Krishna, and 2 more authors
    In The Second Tiny Papers Track at ICLR 2024, 2024