PhD thesis

  1. New Faithfulness-Centric Interpretability Paradigms for Natural Language Processing
    by , with Siva Reddy and Sarath Chandar as supervisors.
    Polytechnique Montreal ⸺ November 2024.
    [thesis]

Master's thesis

  1. Training Neural Networks to Perform Structured Prediction Task
    by , with Sarath Chandar and Amal Zouaq as supervisors.
    Polytechnique Montréal ⸺ August 2024.
    [thesis]

  2. The Role of Continual Learning and Adaptive Computation in Improving Computational Efficiency of Deep Learning
    by , with Irina Rish and Sarath Chandar as supervisors.
    Université de Montréal ⸺ January 2024.
    [thesis]

  3. Towards Adaptive Deep Model-Based Reinforcement Learning
    by , with Sarath Chandar as supervisor.
    Université de Montréal ⸺ November 2023.
    [thesis]

  4. Combining Reinforcement Learning and Constraint Programming for Sequence-Generation Tasks with Hard Constraints
    by , with Gilles Pesant and Sarath Chandar as supervisors.
    Polytechnique Montréal ⸺ November 2022.
    [thesis]

  5. Benchmarking Bias Mitigation Algorithms in Representation Learning through Fairness Metrics
    by , with Sarath Chandar as supervisor.
    Université de Montréal ⸺ July 2022.
    [thesis]

  6. TAG: Task-based Accumulated Gradients for Lifelong learning
    by , with Balaraman Ravindran and Sarath Chandar as supervisors.
    Indian Institute of Technology, Madras ⸺ January 2022.
    [thesis]

  7. PatchUp: A Feature-Space Block-Level Regularization Technique for Convolutional Neural Networks
    by , with Sarath Chandar as supervisor.
    Université de Montréal ⸺ October 2021.
    [thesis]

  8. IIRC: Incremental Implicitly-Refined Classification
    by , with Sarath Chandar as supervisor.
    Université de Montréal ⸺ May 2021.
    [thesis]

  9. Continuous Coordination As a Realistic Scenario for Lifelong Learning
    by , with Aaron Courville and Sarath Chandar as supervisors.
    Université de Montréal ⸺ April 2021.
    [thesis]