Yash Pandya

Senior Research Engineer

Publications

View by:

- Privileged, but Biased: How PI-Conditioned Teachers Break Self-Distillation
  
  Sarthak Harne, Chinmay Karkar, Yash Pandya, Ahmed Awadallah, Akshay Nambi
  
  August 2026
  
  arXiv | August 2026
  
  Publication
- Echoverse: Deep, Evolving Environments for Training Computer-Use Agents at Scale
  
  Yash Pandya, Sahil Gupta, Sarthak Harne, Archana Yadav, Kavyansh Chourasia, Hussein Mozannar, Vibhav Vineet, Sara Abdali, Corby Rosset, Yash Lara, Ahmed Awadallah, Ece Kamar, Akshay Nambi
  
  July 2026
- Learning When to Act or Refuse: Guarding Agentic Reasoning Models for Safe Multi-Step Tool Use
  
  Aradhye Agarwal, Gurdit Siyan, Yash Pandya, Joykirat Singh, Akshay Nambi, Ahmed Awadallah
  
  2026 International Conference on Machine Learning (ICML) | July 2026
  
  Publication
- Fara-1.5: Scalable Learning Environments for Computer Use Agents
  
  Ahmed Awadallah, Sahil Gupta, Yash Lara, Yadong Lu, Hussein Mozannar, Akshay Nambi, Zach Nussbaum, Yash Pandya, Aravind Rajeswaran, Corby Rosset, Alexey Taymanov, Luiz do Valle, Vibhav Vineet, Spencer Whitehead, Andrew Zhao
  
  June 2026
  
  arXiv | June 2026
  
  Publication
- Exposing Weak Links in Multi-Agent Systems under Adversarial Prompting
  
  Nirmit Arora, Sathvik Joel, Ishan Kavathekar, Palak LNU, Rohan Gandhi, Yash Pandya, Tanuja Ganu, Aditya Kanade , Akshay Nambi
  
  AAMAS 2026 SE Work | May 2026
  
  PDF
- Scaling Agentic Capabilities, Not Context: Efficient Reinforcement Finetuning for Large Toolspaces
  
  Karan Gupta, Pranav Vajreshwari, Yash Pandya, Raghav Magazine, Akshay Nambi, Ahmed Awadallah
  
  ICLR Agents in the Wild (Spotlight) | March 2026
  
  Publication
- Fara-7B: An Efficient Agentic Model for Computer Use
  
  Ahmed Awadallah, Yash Lara, Raghav Magazine, Hussein Mozannar, Akshay Nambi, Yash Pandya, Aravind Rajeswaran, Corby Rosset, Alexey Taymanov, Vibhav Vineet, Spencer Whitehead, Andrew Zhao
  
  MSR-TR-2025-54 | November 2025
  
  Published by Microsoft
  
  PDF Access

- Privileged, but Biased: How PI-Conditioned Teachers Break Self-Distillation
  
  Sarthak Harne, Chinmay Karkar, Yash Pandya, Ahmed Awadallah, Akshay Nambi
  
  August 2026
  
  arXiv | August 2026
  
  Publication
- Echoverse: Deep, Evolving Environments for Training Computer-Use Agents at Scale
  
  Yash Pandya, Sahil Gupta, Sarthak Harne, Archana Yadav, Kavyansh Chourasia, Hussein Mozannar, Vibhav Vineet, Sara Abdali, Corby Rosset, Yash Lara, Ahmed Awadallah, Ece Kamar, Akshay Nambi
  
  July 2026
- Learning When to Act or Refuse: Guarding Agentic Reasoning Models for Safe Multi-Step Tool Use
  
  Aradhye Agarwal, Gurdit Siyan, Yash Pandya, Joykirat Singh, Akshay Nambi, Ahmed Awadallah
  
  2026 International Conference on Machine Learning (ICML) | July 2026
  
  Publication
- Fara-1.5: Scalable Learning Environments for Computer Use Agents
  
  Ahmed Awadallah, Sahil Gupta, Yash Lara, Yadong Lu, Hussein Mozannar, Akshay Nambi, Zach Nussbaum, Yash Pandya, Aravind Rajeswaran, Corby Rosset, Alexey Taymanov, Luiz do Valle, Vibhav Vineet, Spencer Whitehead, Andrew Zhao
  
  June 2026
  
  arXiv | June 2026
  
  Publication
- Exposing Weak Links in Multi-Agent Systems under Adversarial Prompting
  
  Nirmit Arora, Sathvik Joel, Ishan Kavathekar, Palak LNU, Rohan Gandhi, Yash Pandya, Tanuja Ganu, Aditya Kanade , Akshay Nambi
  
  AAMAS 2026 SE Work | May 2026
  
  PDF
- Scaling Agentic Capabilities, Not Context: Efficient Reinforcement Finetuning for Large Toolspaces
  
  Karan Gupta, Pranav Vajreshwari, Yash Pandya, Raghav Magazine, Akshay Nambi, Ahmed Awadallah
  
  ICLR Agents in the Wild (Spotlight) | March 2026
  
  Publication
- Fara-7B: An Efficient Agentic Model for Computer Use
  
  Ahmed Awadallah, Yash Lara, Raghav Magazine, Hussein Mozannar, Akshay Nambi, Yash Pandya, Aravind Rajeswaran, Corby Rosset, Alexey Taymanov, Vibhav Vineet, Spencer Whitehead, Andrew Zhao
  
  MSR-TR-2025-54 | November 2025
  
  Published by Microsoft
  
  PDF Access
- Privileged, but Biased: How PI-Conditioned Teachers Break Self-Distillation
  
  Sarthak Harne, Chinmay Karkar, Yash Pandya, Ahmed Awadallah, Akshay Nambi
  
  August 2026
  
  arXiv | August 2026
  
  Publication
- Exposing Weak Links in Multi-Agent Systems under Adversarial Prompting
  
  Nirmit Arora, Sathvik Joel, Ishan Kavathekar, Palak LNU, Rohan Gandhi, Yash Pandya, Tanuja Ganu, Aditya Kanade , Akshay Nambi
  
  AAMAS 2026 SE Work | May 2026
  
  PDF

- Privileged, but Biased: How PI-Conditioned Teachers Break Self-Distillation
  
  Sarthak Harne, Chinmay Karkar, Yash Pandya, Ahmed Awadallah, Akshay Nambi
  
  August 2026
  
  arXiv | August 2026
  
  Publication
- Fara-1.5: Scalable Learning Environments for Computer Use Agents
  
  Ahmed Awadallah, Sahil Gupta, Yash Lara, Yadong Lu, Hussein Mozannar, Akshay Nambi, Zach Nussbaum, Yash Pandya, Aravind Rajeswaran, Corby Rosset, Alexey Taymanov, Luiz do Valle, Vibhav Vineet, Spencer Whitehead, Andrew Zhao
  
  June 2026
  
  arXiv | June 2026
  
  Publication
- Echoverse: Deep, Evolving Environments for Training Computer-Use Agents at Scale
  
  Yash Pandya, Sahil Gupta, Sarthak Harne, Archana Yadav, Kavyansh Chourasia, Hussein Mozannar, Vibhav Vineet, Sara Abdali, Corby Rosset, Yash Lara, Ahmed Awadallah, Ece Kamar, Akshay Nambi
  
  July 2026
- Learning When to Act or Refuse: Guarding Agentic Reasoning Models for Safe Multi-Step Tool Use
  
  Aradhye Agarwal, Gurdit Siyan, Yash Pandya, Joykirat Singh, Akshay Nambi, Ahmed Awadallah
  
  2026 International Conference on Machine Learning (ICML) | July 2026
  
  Publication
- Exposing Weak Links in Multi-Agent Systems under Adversarial Prompting
  
  Nirmit Arora, Sathvik Joel, Ishan Kavathekar, Palak LNU, Rohan Gandhi, Yash Pandya, Tanuja Ganu, Aditya Kanade , Akshay Nambi
  
  AAMAS 2026 SE Work | May 2026
  
  PDF
- Scaling Agentic Capabilities, Not Context: Efficient Reinforcement Finetuning for Large Toolspaces
  
  Karan Gupta, Pranav Vajreshwari, Yash Pandya, Raghav Magazine, Akshay Nambi, Ahmed Awadallah
  
  ICLR Agents in the Wild (Spotlight) | March 2026
  
  Publication
- Fara-7B: An Efficient Agentic Model for Computer Use
  
  Ahmed Awadallah, Yash Lara, Raghav Magazine, Hussein Mozannar, Akshay Nambi, Yash Pandya, Aravind Rajeswaran, Corby Rosset, Alexey Taymanov, Vibhav Vineet, Spencer Whitehead, Andrew Zhao
  
  MSR-TR-2025-54 | November 2025
  
  Published by Microsoft
  
  PDF Access

Yash Pandya

Publications

Privileged, but Biased: How PI-Conditioned Teachers Break Self-Distillation

Echoverse: Deep, Evolving Environments for Training Computer-Use Agents at Scale

Learning When to Act or Refuse: Guarding Agentic Reasoning Models for Safe Multi-Step Tool Use

Fara-1.5: Scalable Learning Environments for Computer Use Agents

Exposing Weak Links in Multi-Agent Systems under Adversarial Prompting

Scaling Agentic Capabilities, Not Context: Efficient Reinforcement Finetuning for Large Toolspaces

Fara-7B: An Efficient Agentic Model for Computer Use

Privileged, but Biased: How PI-Conditioned Teachers Break Self-Distillation

Echoverse: Deep, Evolving Environments for Training Computer-Use Agents at Scale

Learning When to Act or Refuse: Guarding Agentic Reasoning Models for Safe Multi-Step Tool Use

Fara-1.5: Scalable Learning Environments for Computer Use Agents

Exposing Weak Links in Multi-Agent Systems under Adversarial Prompting

Scaling Agentic Capabilities, Not Context: Efficient Reinforcement Finetuning for Large Toolspaces

Fara-7B: An Efficient Agentic Model for Computer Use

Privileged, but Biased: How PI-Conditioned Teachers Break Self-Distillation

Exposing Weak Links in Multi-Agent Systems under Adversarial Prompting

Privileged, but Biased: How PI-Conditioned Teachers Break Self-Distillation

Fara-1.5: Scalable Learning Environments for Computer Use Agents

Echoverse: Deep, Evolving Environments for Training Computer-Use Agents at Scale

Learning When to Act or Refuse: Guarding Agentic Reasoning Models for Safe Multi-Step Tool Use

Exposing Weak Links in Multi-Agent Systems under Adversarial Prompting

Scaling Agentic Capabilities, Not Context: Efficient Reinforcement Finetuning for Large Toolspaces

Fara-7B: An Efficient Agentic Model for Computer Use

Contact Yash Pandya

Microsoft Research Lab – India