Ravi Tej Akella

ravitej.akella@gmail.com

About Me

I train end-to-end self-driving neural networks at Tesla AI. Previously, I studied robotics at Carnegie Mellon University, where I was fortunate to work with Prof. Ben Eysenbach, Prof. Ruslan Salakhutdinov, and Prof. Jeff Schneider. Broadly speaking, I am excited about building generally intelligent agents that make our lives easier.

While at CMU, I worked on goal-conditioned reinforcement learning (RL), offline RL, and autonomous navigation in off-road environments (DARPA RACER project). Around this time, I also interned as an ML engineer at Cruise. Before joining CMU, I was a research intern in TensorLab (Caltech), where I closely collaborated with Prof. Kamyar Azizzadenesheli (Purdue University), Prof. Anima Anandkumar (Caltech), Prof. Yisong Yue (Caltech), and Dr. Mohammad Ghavamzadeh (Google Research). I received my Bachelor's degree from IIT Roorkee, with a major in Electronics and Communication Engineering and minors in Computer Science.


B.Tech ECE IIT Roorkee Jul 2014 - May 2018	Research Intern Caltech Oct 2018 - Dec 2020	MS Robotics Carnegie Mellon University Aug 2021 - Aug 2023	ML Engineer Intern Cruise May 2022 - Aug 2022	ML Scientist Tesla Aug 2023 - Present

Publications

Reasoning with Latent Diffusion in Offline Reinforcement Learning

Siddarth Venkatraman*, Shivesh Khaitan*, Ravi Tej Akella*, John Dolan, Jeff Schneider, Glen Berseth

Presented at International Conference on Learning Representations (ICLR) 2024

PDF BibTeX

  @inproceedings{
    venkatraman2024reasoning,
    title={Reasoning with Latent Diffusion in Offline Reinforcement Learning},
    author={Siddarth Venkatraman and Shivesh Khaitan and Ravi Tej Akella and John Dolan and Jeff Schneider and Glen Berseth},
    booktitle={The Twelfth International Conference on Learning Representations},
    year={2024},
    url={https://openreview.net/forum?id=tGQirjzddO}
    }

Distributional Distance Classifiers for Goal-Conditioned Reinforcement Learning

Ravi Tej Akella, Benjamin Eysenbach, Jeff Schneider, Ruslan Salakhutdinov

Presented at New Frontiers in Learning, Control, and Dynamical Systems, ICML Workshop 2023

PDF BibTeX

  @mastersthesis{Akella-2023-136987,
    author = {Ravi Tej Akella},
    title = {Distributional Distance Classifiers for Goal-Conditioned Reinforcement Learning},
    year = {2023},
    month = {July},
    school = {Carnegie Mellon University},
    address = {Pittsburgh, PA},
    number = {CMU-RI-TR-23-27},
    keywords = {Reinforcement Learning, Goal-Conditioned, Dynamical Distance Learning, Contrastive Learning},
    }

Deep Bayesian Quadrature Policy Optimization

Ravi Tej Akella, Kamyar Azizzadenesheli, Mohammad Ghavamzadeh, Anima Anandkumar, Yisong Yue

Proceedings of the 35th AAAI Conference on Artificial Intelligence 2021

Presented at the NeurIPS Deep RL & Real-World RL Workshops 2020

Video Project PDF Poster BibTeX Blog Slides Code

@article{ravi2020DBQPG,
title={Deep Bayesian Quadrature Policy Optimization},
author={Ravi Tej Akella and Kamyar Azizzadenesheli and Mohammad Ghavamzadeh and Anima Anandkumar and Yisong Yue},
journal={arXiv preprint arXiv:2006.15637},
year={2020}
}

Reinforced Multi-task Approach for Multi-hop Question Generation

Deepak Gupta, Hardik Chauhan, Ravi Tej Akella, Asif Ekbal, Pushpak Bhattacharyya

Proceedings of the 28th International Conference on Computational Linguistics (COLING) 2020 (Long Paper)

PDF BibTeX

@inproceedings{gupta2020reinforced,
    title = "Reinforced Multi-task Approach for Multi-hop Question Generation",
    author = "Gupta, Deepak  and
      Chauhan, Hardik  and
      Akella, Ravi Tej  and
      Ekbal, Asif  and
      Bhattacharyya, Pushpak",
    booktitle = "Proceedings of the 28th International Conference on Computational Linguistics",
    month = dec,
    year = "2020",
    address = "Barcelona, Spain (Online)",
    publisher = "International Committee on Computational Linguistics",
    url = "https://www.aclweb.org/anthology/2020.coling-main.249",
    pages = "2760--2775",
}

Enhancing Perceptual Loss with Adversarial Feature Matching for Super-Resolution

Ravi Tej Akella, Shirsendu Sukanta Halder, Arunav Shandilya, Vinod Pankajakshan

Proceedings of the 31st International Joint Conference on Neural Networks (IJCNN) 2020 (Oral)

PDF BibTeX

@INPROCEEDINGS{ravi2020EnhancedPerceptualLoss,
  author={A. R. {Tej} and S. {Sukanta Halder} and A. P. {Shandeelya} and V. {Pankajakshan}},
  booktitle={2020 International Joint Conference on Neural Networks (IJCNN)}, 
  title={Enhancing Perceptual Loss with Adversarial Feature Matching for Super-Resolution}, 
  year={2020},
  volume={},
  number={},
  pages={1-8},
  doi={10.1109/IJCNN48605.2020.9207102}}

A Randomized Kernel-Based Secret Image Sharing Scheme

Ravi Tej Akella, Rekula Raviteja, Vinod Pankajakshan

Presented at the IEEE International Workshop on Information Forensics and Security (WIFS) 2018

PDF BibTeX

@INPROCEEDINGS{ravi2020secretsharing,
  author={A. R. {Tej} and R. {Raviteja} and V. {Pankajakshan}},
  booktitle={2018 IEEE International Workshop on Information Forensics and Security (WIFS)}, 
  title={A Randomized Kernel-Based Secret Image Sharing Scheme}, 
  year={2018},
  volume={},
  number={},
  pages={1-7},}

Latest Updates

[Apr 2024] "Reasoning with Latent Diffusion in Offline Reinforcement Learning" accepted to ICLR 2024.
[Sept 2023] Started as a Machine Learning Scientist at Tesla AI.
[Jul 2023] "Distributional Distance Classifiers for Goal-Conditioned Reinforcement Learning" accepted to the New Frontiers in Learning, Control, and Dynamical Systems Workshop, ICML 2023.
[May 2022] Starting as an ML Engineer Intern in the Maneuver Planning team at Cruise.
[Aug 2021] Starting my M.S. Robotics (MSR) degree at the Robotics Institute, Carnegie Mellon University (CMU).
[Aug 2021] Attended Machine Learning Summer School (MLSS) Taipei.
[Dec 2020] "Deep Bayesian Quadrature Policy Optimization" accepted to AAAI 2021.
[Oct 2020] "Deep Bayesian Quadrature Policy Optimization" accepted to NeurIPS Deep RL and Real-World RL workshops.
[Sep 2020] "Reinforced Multi-task Approach for Multi-hop Question Generation" accepted to COLING 2020.
[May 2020] " Enhancing Perceptual Loss with Adversarial Feature Matching for Super-Resolution" accepted to IJCNN 2020.
[Oct 2019] Serving as Secondary Reviewer for AAAI 2020.
[Sept 2019] Serving as Secondary Reviewer for Machine Learning and the Physical Sciences (ML4PS), NeurIPS (2019) workshop.

Ravi Tej Akella

ravitej.akella@gmail.com

About Me

Publications

Recent Posts

Bayesian Quadrature for Policy Gradient

Latest Updates