Dilith Jayakody

PhD student, Computer Science at Dalhousie University. Researching Meta-Reinforcement Learning with Prof. Janarthanan Rajendran.

I work on how reinforcement learning agents can learn their own intrinsic rewards to become more sample-efficient. Before Dal, I completed my BSc in Computer Science and Engineering at the University of Moratuwa and lectured there on a contract basis. I’m broadly interested in RL, computer vision, and the math that holds modern AI together.

News

  • May 2026 Started this academic homepage — see the new publications and blog pages.
  • 2024 Paper accepted at WMT 2024: "Back to the Stats: Rescuing Low Resource Neural Machine Translation with Statistical Methods."
  • 2024 Paper accepted at ACL 2024: "Shoulders of Giants: A Look at the Degree and Utility of Openness in NLP Research."
  • 2023 Started PhD in Computer Science at Dalhousie University with Prof. Janarthanan Rajendran, working on meta-reinforcement learning.

Selected publications

  • Back to the Stats: Rescuing Low Resource Neural Machine Translation with Statistical Methods
    V. Menan, D. Jayakody, N. de Silva, A. Fernando, S. Ranathunga
    Proceedings of the Ninth Conference on Machine Translation (WMT), 2024.
  • Shoulders of Giants: A Look at the Degree and Utility of Openness in NLP Research
    S. Ranathunga, N. de Silva, D. Jayakody, A. Fernando
    Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (ACL), 2024.
  • Few-shot Multispectral Segmentation with Representations Generated by Reinforcement Learning
    D. Jayakody, T. Ambegoda
    arXiv preprint, 2023.
    arXiv

All publications →

Recent writing

Kalman Filters - A Quick Introduction

Kalman Filters - A Quick Introduction

Sharpness-Aware Minimization (SAM) - A Quick Introduction

Sharpness-Aware Minimization (SAM) - A Quick Introduction

Model Predictive Path Integral (MPPI) - A Quick Introduction

Model Predictive Path Integral (MPPI) - A Quick Introduction

The Reparameterization Trick - Clearly Explained

The Reparameterization Trick - Clearly Explained

Model-based vs. Model-free Reinforcement Learning - Clearly Explained

Model-based vs. Model-free Reinforcement Learning - Clearly Explained

Q-Learning and SARSA in RL - Similarities and Differences Explained

Q-Learning and SARSA in RL - Similarities and Differences Explained

All posts →