This is the website of Sam Griesemer. I'm a 4th year CS PhD student at USC, advised by Professor Yan Liu. Prior to USC, I studied computer science and
probability/statistics at Washington University, and worked as a research intern at the MIT Lincoln Lab, focusing on deep learning for simulation-based inference.
You can reach me at samgriesemer [at] gmail [dot] com. You can also contact me on
Twitter or find some of my code on Github. For more about me and this
site, see the about page. For more about what I'm
currently working on, see now.
Latest news
Mar '25 | Find our ICLR '25 Frontiers in Probabilistic Inference paper here.
Atomic Posterior Ensembles for Simulation-Based InferenceSam Griesemer, Willie Neiswanger, Yan Liu
Dec '24 | Our work on active SBI has been accepted to NeurIPS '24. Find the manuscript on arXiv here.
Active Sequential Posterior Estimation for Sample-Efficient Simulation-Based InferenceSam Griesemer, Defu Cao, Zijun Cui, Carolina Osorio, Yan Liu
You can also find our code (seqinf Python package) here.
Feed
An Architecture for Deep, Hierarchical Generative Models Neural Geometric Level of Detail Switch Transformers Github1s paint.wtf WordsEye
Blog
Introduction to basic probability
An overview of fundamental probability conceptsRandom sampling without replacement
A quick runtime analysis of methods for taking random samples without replacementSetting up a remote jupyter notebook for a website
In-depth tutorial for creating a Jupyter lab accessible from your domain
Projects
Panja
Super simple static site generator using [Pan]doc and Jin[ja], with added note processing utilitiesproblib ui
Frontend component to problib app pipelineProgramming templates
Collection of templates for a number of programming environments
Research
Image Quilting for Texture Synthesis
Implementing texture synthesis algorithms from (Efros & Freeman, 2001)Learning Across Repeated Congestion Games
Observing simulated agent behaviour under Braess’s paradoxDetection and Analysis of Subreddit Communities
Graph theoretic analysis of over 40GB of Reddit data