Avatar

Skanda Vaidyanath

Founding AI Researcher

Yutori

About

I am a Founding AI Researcher at Yutori. We’re building multi-modal computer use agents.

Our latest model, n2, is a 27B model at the cost-performance frontier for CUA models and is SOTA on four of five evaluated benchmarks. Check out the blog here.

My current research focuses on training LLMs with reinforcement learning to solve challenging, long-horizon, multi-step tasks. These tasks often provide only sparse, binary feedback, creating difficult credit-assignment and exploration challenges. To address these challenges, I’m exploring techniques that enable models to learn from natural-language feedback. I’m also eager to explore applications of LLMs to the natural sciences and build systems capable of conducting frontier research.

If you are interested in my work or would like to chat about technical interests we might share, feel free to get in touch!

Interests

  • Reinforcement Learning
  • LLMs

Education

  • MS in Computer Science, AI track, 2023

    Stanford University

  • BE (Hons) in Computer Science with a Minor in Data science, 2020

    BITS Pilani, Hyderabad Campus

Blog Posts

An Overview of Bandits

In this post, we will take a look at the bandit problem and discuss some solution strategies. This is a fairly introductory overview so …
Jan 21, 2020 13 min read Reinforcement Learning

Inverse Reinforcement Learning

This is a review of the paper Algorithms for Inverse Reinforcement Learning. I recommend some reinforcement learning (RL) basics before …
Jan 19, 2020 9 min read Reinforcement Learning

Modeling RL Problems

Disclaimer: The content for this article does not come from any textbook or other reliable sources. They are observations made purely …
Jan 16, 2020 9 min read Reinforcement Learning

Bridging the Gaps With Reinforcement Learning

In this post, I will be talking about a unique way to use reinforcement learning (RL) in deep learning applications. I definitely …
Nov 4, 2019 7 min read Reinforcement Learning, Deep Learning