Adam Jozefiak

Adam Jozefiak

PhD candidate, Operations Research Center, MIT

I am a PhD candidate in the Operations Research Center at MIT, advised by Prof. Vivek F. Farias. My work is on efficient inference for generative models, reinforcement learning, and continual learning.

I work closely with industry research collaborators, including Anheuser-Busch InBev, KNAPP, Meta, and Takeda, on deploying these methods in practice. I am currently working with the Minerals Stewardship Consortium (Rio Tinto, Vale, and BHP) on generative models for orebody knowledge and reinforcement learning for mine planning.

In summer 2026 I was an applied scientist intern at Amazon, where I used reinforcement learning to train replenishment policies for its regionalized distribution and fulfillment network.

Before MIT I completed an M.Sc. in Computer Science and a B.Sc. in Computer Science and Mathematics at the University of British Columbia, where I was advised by Prof. Bruce Shepherd and worked on approximation algorithms for combinatorial optimization. I also worked with Prof. Jesse Perla at the Vancouver School of Economics on computational macroeconomics.

Email
jozefiak (at) mit (dot) edu

Research

  1. The Value of Covariance Matching in Gaussian DDPMs and the Lanczos Sampler

    Md Sahil Akhtar, Aymane El Gadarri, Vivek F. Farias, Adam D. Jozefiak

    Preprint, 2026.

  2. OS-Pruner: Pruning Chains-of-Thought of Reasoning Models via Optimal Stopping

    Mohammed Ehab, Aymane El Gadarri, Vivek F. Farias, Adam D. Jozefiak, Ciamac C. Moallemi

    Preprint, 2026.

  3. Self-Normalized Resets for Plasticity in Continual Learning

    Vivek F. Farias, Adam D. Jozefiak

    Preliminary version: ICLR 2025.

    Extended version: major revisions, Operations Research.

  4. Diversity Embeddings and the Hypergraph Sparsest Cut

    Adam D. Jozefiak, F. Bruce Shepherd

    Minor revision, Discrete Mathematics & Theoretical Computer Science.

  5. A Knapsack Intersection Hierarchy

    Adam D. Jozefiak, F. Bruce Shepherd, Noah Weninger

    Operations Research Letters, 51(1):72–78, 2023.

  6. Eigenvalues and Eigenvectors of Tau Matrices with Applications to Markov Processes and Economics

    Sven-Erik Ekström, Carlo Garoni, Adam D. Jozefiak, Jesse Perla

    Linear Algebra and its Applications, 627:41–71, 2021.

Teaching