Tristan Saidi

Tristan Luca Saidi

I am a PhD student at Columbia University in the department of Computer Science (on leave). I'm fortunate to be working with Professor Andrew J. Blumberg on theoretically grounded methods that use discrete graph curvature to improve geometric data analysis.

During the first year of my PhD I had the privilege of conducting research in Professor Matei Ciocarlie's Robotic Manipulation and Mobility (ROAM) Lab . I primarily worked on robot learning for dexterous manipulation, and spent some time working on reinforcement learning for snake locomotion as well. Prior to my time with robotics, I spent a year doing research for Professor Kenneth Shepard in the Bioelectronic Systems Lab.

Email / CV / Github

Research

I'm interested in developing theoretically justified approaches for recovering the geometric structure of real world data. Primarily, I'm interested in improving manifold learning techniques for the benefit of biologists working with single-cell RNA sequencing data.

	Recovering Manifold Structure Using Ollivier-Ricci Curvature Tristan Luca Saidi, Abigail Hickok, Andrew J. Blumberg Submitted to the Thirteenth International Conference on Learning Representations (ICLR), 2024 arXiv / code
	Diffusion Models Are Promising for Ab Initio Structure Solutions from Nanocrystalline Powder Diffraction Data Gabe Guo, Tristan Luca Saidi, Maxwell Terban, Simon J.L. Billinge Hod Lipson Submitted to Nature Materials, 2024 arXiv / code
	R×R: Rapid eXploration for Reinforcement Learning via Sampling-based Reset Distributions and Imitation Pre-training Gagan Khandate, Tristan Luca Saidi, Siqi Shang, Eric Chang, Johnson Adams, Matei Ciocarlie RSS Special Issue: Autonomous Robots*, 2024 arXiv / Springer
	Sampling Based Exploration for Reinforcement Learning of Dexterous Manipulation Gagan Khandate, Siqi Shang, Eric Chang, Tristan Saidi, Yang Liu, Seth Dennis, Johnson Adams, Matei Ciocarlie Robotics: Science and Systems (RSS), 2023 project page / arXiv /

Projects

Some projects I have been working on recently in classes, research or my free time!

	Exploring Adversarial Perturbations to Diffusion Models Tristan Saidi, Leon Zhou STCS 6701: Foundations of Graphical Models, Fall 2023 code / writeup /
	Examining the geometry of neural mode connecting loss subspaces Ting Chen, Tristan Saidi, COMS 4995: Geometric Data Analysis, Spring 2023 code / writeup /
	On the Effect of Unsupervised Regularization for Image Classification Tristan Saidi, Sagarika Sharma, Mitali Juneja, Nikhilesh Belulkar COMS 4995: Neural Networks and Deep Learning, Fall 2022 code / writeup /
	Recreation of Atari's River Raid Yongmao Luo, Tristan Saidi, Zhaomeng Wang, Jakob Steins CSEE 4840: Embedded Systems, Spring 2022 code /
	Verilog Implementation of an Finite Impulse Response (FIR) filter Tristan Saidi, Bradley Sears CSEE 4823: Advanced Logic Design, Fall 2021 code /

This website template came from Jon Barron! Check out the source code.