Marko Sterbentz

Computer Scientist,Programmer,AI Enthusiast

A little about me

I am a computer scientist with a passion for artificial intelligence and natural language processing. I'm currently pursuing a Ph.D. in computer science at Northwestern University where I research narrative generation, commonsense reasoning, and explainable AI methods.

Research Interests

  • Narrative Generation: Building systems that track the state of the world, the agents operating in this environment, and producing a satisfying story in natural language

  • Commonsense Reasoning: Modeling why people act in the ways they do, determining what impact this has on those around them, and integrating this knowledge with the reasoning capabilities of artificial agents

  • Explainable AI: Enabling artificial intelligence systems to provide reasonable answers into their decision making process and facilitating better human-agent interaction

Publications

[1] Money, James H., Marko Sterbentz, Nathan Morrical, Thomas Szewczyk, and Landon Woolley. "GPGPU Enabled Ray Directed Adaptive Volume Visualization for High Density Scans." In Proceedings of the Practice and Experience on Advanced Research Computing, p. 62. ACM, 2018. [paper]

Education and Experience






  • Doctor of Philosophy

    Computer Science - In Progress

    Master of Science

    Computer Science - 2019

    Bachelor of Science

    Computer Science - 2017

Lawrence Livermore National Laboratory

I developed a new software component with C++ and Python for performing material interface reconstruction in LLNL's open-source HPC framework Axom in order to improve software development efforts for large-scale simulations. This project is open source and the code is available on GitHub.


Poster Source Code
Idaho National Laboratory

Developed improvements to a large-scale volume visualization system with C#, HLSL, and compute shaders. Utilized ray-marching in tandem with a specialized data format to enable interactive visualization of exascale data in immersive environments. Presented associated research paper at PEARC18 conference in July 2018.


Paper Source Code
Undergraduate Research

Wrote software in C++ that interfaced with a LiDAR sensor and an inertial measurement unit (IMU) to extract the data, send it over a wireless network, and accurately recreate the scanned environment on the user’s laptop in real time. Work performed as part of a study to determine the state of plant life in Idaho utilizing unmanned aerial vehicle LiDAR data.


Poster Source Code

Contact

If you would like to contact me or collaborate on a project with me, please reach out to me at markosterbentz@u.northwestern.edu or stop by my office in Mudd Hall 3209.