Home Research Talks Interns

Harkirat Behl


profile-pic



Member of Technical Staff
Physics of AGI, managed by Sébastien Bubeck
Microsoft AI

Education
PhD, University of Oxford, 2021       Advisors: Phil Torr, Pawan Kumar
BTech, Indian Institute of Technology, Kanpur, 2018

 

Research

[New] Phi-3: A Highly Capable Language Model Locally on Your Phone
Marah Abdin, Sam Ade Jacobs, Ammar Ahmad Awan, Jyoti Aneja, Ahmed Awadallah, Hany Awadalla, Nguyen Bach, Amit Bahree, Arash Bakhtiari, Harkirat Behl, Alon Benhaim, Misha Bilenko, Johan Bjorck, Sébastien Bubeck, ....
  Hugging Face's logo 17,207,626  
Phi-3-mini   Phi-3-small   Phi-3-medium   Phi-3-vision  
In Popular Media  

Phi-2: The surprising power of small language models
Marah Abdin, Jyoti Aneja, Sebastien Bubeck, Caio César Teodoro Mendes, Weizhu Chen, Allie Del Giorno, Ronen Eldan, Sivakanth Gopi, Suriya Gunasekar, Mojan Javaheripi, Piero Kauffmann, Yin Tat Lee, Yuanzhi Li, Anh Nguyen, Gustavo de Rosa, Olli Saarikivi, Adil Salim, Shital Shah, Michael Santacroce, Harkirat Singh Behl, Adam Taumann Kalai, Xin Wang, Rachel Ward, Philipp Witte, Cyril Zhang, Yi Zhang
Microsoft Research Blog 2023
Hugging Face's logo 4,831,618  


    TOI logo     Silicon  
  The VergeThe Verge logo.   Silicon  
 
 

Peekaboo: Interactive Video Generation via Masked-Diffusion
Yash Jain*, Anshul Nasery*, Vibhav Vineet, Harkirat Behl
CVPR 2024
     

TOI logo  

Dataset-aware Mixture-of-Experts
Yash Jain, Harkirat Behl, Zsolt Kira, Vibhav Vineet
NeurIPS 2023
   

Textbooks Are All You Need
Suriya Gunasekar, Yi Zhang, Jyoti Aneja, Caio Cesar Teodoro Mendes, Allie Del Giorno, Sivakanth Gopi, Mojan Javaheripi, Piero Kauffmann, Gustavo de Rosa, Olli Saarikivi, Adil Salim, Shital Shah, Harkirat Singh Behl, Xin Wang, Sebastien Bubeck, Ronen Eldan, Adam Tauman Kalai, Yin Tat Lee, Yuanzhi Li
arXiv 2023
  Hugging Face's logo 132,653  

Synced
YouTube
224,000 views
 

Efficiently Robustify Pre-Trained Models
Nishant Jain, Harkirat Behl, Yogesh Rawat, Vibhav Vineet
ICCV 2023
 

X-Decoder: Generalized Decoding for Pixel, Image and Language
Xueyan Zou* , Zi-Yi Dou*, Jianwei Yang*, Zhe Gan, Linjie Li, Chunyuan Li, Xiyang Dai, Harkirat Behl, Jianfeng Wang, Lu Yuan, Nanyun Peng, Lijuan Wang, Yong Jae Lee, Jianfeng Gao
CVPR 2023
       

 

Neural-Sim: Learning to Generate Training Data with NeRF
Yunhao Ge, Harkirat Behl, Jiashu Xu, Suriya Gunasekar, Neel Joshi, Yale Song, Xin Wang, Laurent Itti, and Vibhav Vineet
ECCV 2022
   

Automated and Verified Deep Learning
Harkirat Behl
University of Oxford, 2021
 

Overcoming the Convex Barrier for Simplex Inputs
Harkirat Behl, M. Pawan Kumar, Philip Torr, Krishnamurthy (Dj) Dvijotham
NeurIPS 2021
   

Scaling the Convex Barrier with Sparse Dual Solvers
Alessandro De Palma, Harkirat Behl, Rudy Bunel, Philip Torr, M. Pawan Kumar
JMLR 2024
   

Scaling the Convex Barrier with Active Sets
Harkirat Singh Behl*, Alessandro De Palma*, Rudy Bunel, Philip Torr, M. Pawan Kumar
ICLR 2021
     

Progressive Skeletonization: Trimming more fat from a network at initialization
Pau Jorge, Amartya Sanyal, Harkirat Singh Behl, Philip Torr, Gregory Rogez, Puneet Dokania
ICLR 2021
     

STEER: Simple Temporal Regularization For Neural ODEs
Arnab Ghosh, Harkirat Singh Behl, Emilien Dupont, Philip Torr, Vinay Namboodiri
NeurIPS 2020
   

AutoSimulate: (Quickly) Learning Synthetic Data Generation
Harkirat Singh Behl, Atılım Güneş Baydin, Ran Gal, Philip Torr, Vibhav Vineet
ECCV 2020
     

Meta-Learning Deep Visual Words for Fast Video Object Segmentation
Harkirat Singh Behl, Mohammad Najafi, Anurag Arnab, Philip Torr
IROS 2020 (Oral)
NeurIPS 2019 W (Oral)
   

 

Alpha MAML: Adaptive Model-Agnostic Meta-Learning
Harkirat Singh Behl, Atılım Güneş Baydin, Philip Torr
ICML 2019 W
     

Incremental Tube Construction for Human Action Detection
Harkirat Singh Behl, Michael Sapienza, Gurkirt Singh, Suman Saha, Fabio Cuzzolin, Philip Torr
BMVC 2018
   

 

Recent Talks

The Small Language Models Revolution
  • Microsoft IC3-AI Team Talk, Oct. 2024
  • Microsoft Reverb, Sep. 2024
    (offsite of 100 Partner+ Microsoft leaders, held once a year)
  • HMG Tech Intelligence team, June 2024
  • Microsoft AI Compilers Workshop, May 2024
  • Github, AI for Code Reading Group, May 2024
  • Microsoft PoAA (Pals of Autonomous Agents), May 2024
Phi-2: The surprising power of small language models
  • JEM council meeting, Apr 2024
    (Microsoft E+D research leaders meeting)
  • Perception Cloud team, Zurich, Mar 2024
Phi-bonnaci: AI Math Tutor
  • Microsoft AI Frontiers Offsite, Jan 2024
  • Bill & Melinda Gates Foundation Representatives, Dec 2023

Interns