Vaish Shrivastava

I am a Senior Researcher at Microsoft Research in Redmond, where I work on reinforcement learning, reasoning, and computer-use agents.

I earned my Master's in Computer Science at Stanford University, advised by Percy Liang and was a student researcher at the Allen Institute for AI.

Previously, I was an Applied Scientist at Microsoft AI, working on efficient and personalized language models. I received my Bachelor's in Computer Science from Caltech.

Vaish Shrivastava

Research

I'm interested in language models and agents that can learn from interaction, reason reliably, and use computation efficiently. My current work explores reinforcement learning for computer-use agents, learning world models from environment feedback, and reducing unnecessary reasoning. I also study how to estimate model confidence, improve consistency, and understand safety failures.

Publications

Phi-4-reasoning and Phi-4-reasoning-plus performance on reasoning benchmarks

Phi-4-reasoning Technical Report

Marah Abdin, Sahaj Agarwal, Ahmed Awadallah, Vidhisha Balachandran, Harkirat Behl, Lingjiao Chen, Gustavo de Rosa, Suriya Gunasekar, Mojan Javaheripi, Neel Joshi, Piero Kauffmann, Yash Lara, Caio César Teodoro Mendes, Arindam Mitra, Besmira Nushi, Dimitris Papailiopoulos, Olli Saarikivi, Shital Shah, Vaishnavi Shrivastava, Vibhav Vineet, Yue Wu, Safoora Yousefi, Guoqing Zheng

Authors listed alphabetically by last name.

arXiv 2025

arXiv