Jake Frischmann
Discovering and solving problems hidden in software, learning, and physical systems.
- 2026
LEAD Lab language environments Undergraduate research assistant - 2026
VERTEX rover simulation Undergraduate researcher - 2025
SpiRobControls (Drexel Zheng Lab) Research intern - 2024
BlokusDuo ML Agents ML engineer and research author - Rare-cell recognition and quantum advantage Exploratory project: classical clustering baselines for rare-cell recognition, laid out as the starting point for testing a quantum advantage.
- Why D&D Feels Like Systems Design After years of playing and DMing, D&D has become one of my favorite ways to think about rules, randomness, and balance.
- Why Orbit UMD Has a Demo Mode Building a real demo path forced cleaner data boundaries, smoother walkthroughs, and better feedback from faculty and staff.
- What I Learned Building PPO + MCTS for Blokus Duo A practical writeup on training self-play agents, evaluating them honestly, and figuring out where search actually helped.
- What First-Time Hackathon Teams Usually Need Patterns from mentoring beginner teams at Bitcamp, mostly around bad scope decisions, silent blockers, and getting one thing working end to end.
-
Full CV I am open to PhD programs, research labs, and collaborations where careful implementation matters as much as the question. Current work spans children's language environments, planetary rover simulation and control, soft-robotics control, reinforcement learning, and education software. - One-page resume I study how computation, physics, and biological data can explain learning and decision-making. Dual B.S. expected May 2028; accelerated M.S. in Computer Science expected May 2029; planning Ph.D. applications in computational science.
- Publications Frischmann, J., et al. BlokusDuo deep reinforcement learning agents: architecture and experiments. PGSS Journal, 2024. Peer-reviewed; 20 pages. Main author.
- University of Maryland, College Park B.S. Computer Science and B.S. Applied Physics (dual degree). Minor in Quantum Science and Engineering. Accelerated B.S./M.S. in Computer Science (M.S. May 2029).
- Honors GPA 3.85; Dean’s List 2025, 2026
- Pennsylvania Governor’s School for the Sciences (PGSS) Led a BlokusDuo agents project (minimax, MCTS, MCTS with a CNN, PPO) in a residential program with a 7% admit rate; MCTS with a CNN won 66% as player 1 against standard MCTS and consistently beat experienced human players.
-
About me I am Jake Frischmann, a dual-degree student in Computer Science and Applied Physics at the University of Maryland, with a minor in Quantum Science and Engineering and accelerated CS master's coursework in progress (M.S. expected May 2029). - Get in touch Feel free to reach out about research, projects, or anything you found intriguing.
- Leadership and service I have been a Best Buddies buddy since high school, volunteer with KEEN Greater DC, lead software on the Henson Day iOS app in AppDev, and organize the quantum track at Bitcamp.
- Soccer I captain an intramural team at Maryland. The rest of the week is BJJ, climbing, swimming, and strength training.
- D&D Ten years as a player, three as a dungeon master. I treat the rulebook as a design space, not a constraint.
- Music and the rest Music, sci-fi, board games, and collecting playing cards. I also run a fantasy book club on campus, which is where most of the dungeon-master notes get tested.