Vikas Yadav
Staff Research Scientist at Meta
Menlo Park, California, United States
Summary
👤
Senior🎓
Top SchoolVikas Yadav is a Staff Research Scientist with 12 years of experience building and fine-tuning large language models and latency-sensitive ML systems, currently at Meta Superintelligence Labs after leading post-training and reasoning work on Google’s Gemini family. He pioneered novel training recipes and RLHF pipelines that materially improved reasoning, factuality, coding and instruction-following—contributions credited with state-of-the-art gains versus contemporaries like GPT-4 on several benchmarks. Prior roles span Google Assistant, Wear OS and Azure, where he shipped production models, cut latency, and engineered robust replication and backup systems, demonstrating a rare blend of research rigor and product-grade engineering. Based in Menlo Park, he combines deep model-level expertise with practical data collection and user-signal extraction techniques to align models to real-world user preferences. An IIT Roorkee alumnus, he pairs competitive programming roots with systems-level optimizations that accelerate deployment of advanced AI features.
12 years of coding experience
9 years of employment as a software developer
Indian Institute of Technology Roorkee
Github Skills (8)
ddpg7
deep-reinforcement-learning6
pytorch6
reinforcement-learning6
deep-learning2
machine-learning1
python1
data-science1
Programming languages (3)
Github contributions (5)
vy007vikas/PyTorch-ActorCriticRL
Aug 2017 - Sep 2017
PyTorch implementation of DDPG algorithm for continuous action reinforcement learning problem.
Contributions:21 commits in 14 days
pytorchddpgddpg-algorithmreinforcement-learningcontinuous
vy007vikas/ProNote
Jan 2014 - May 2014
Contributions:10 commits in 3 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.