Sam Wang
I am a student studying AI and Robotics at the University of Pennsylvania. I am also a member of the Perception Action & Learning Group of the Penn's GRASP Laboratory.
This past year, I was a student researcher at the Allen Institute for AI on the PRIOR and Robotics teams, advised by Prof. Ranjay Krishna, Dr. Jiafei Duan, and Prof. Dieter Fox, where I contributed to the open MolmoAct 2 robotics foundation model.
Previously, in Summer 2025, I interned on Dyna Robotics' research team, where I trained frontier VLA models for real world deployment. At Penn, I have worked under the guidance of Dr. Jason Ma, Prof. Dinesh Jayaraman, Prof. Osbert Bastani, alongside Will Liang and Johnny Wang.
I've coauthored three papers on robot learning, which were published at ICLR, CoRL and RSS. My previous research on applying deep learning for rapid cardiac screening earned a Second Grand Prize in Robotics and Intelligent Machines at ISEF 2023.
Github
Google Scholar
LinkedIn
Twitter
MolmoAct 2: An open foundation for robots that work in the real world
Haoquan Fang*§, Jiafei Duan*§, Donovan Clay§, Sam Wang§, Shuo Liu§, Weikai Huang§, Xiang Fan§, Wei-Chuan Tsai§, Shirui Chen§, Yi Ru Wang§, Shanli Xing§, Jaemin Cho, Jae Sung Park, Ainaz Eftekhar, Peter Sushko, Karen Farley, Angad Wadhwa, Cole Harrison, Winson Han, Ying-Chun Lee, Eli VanderBilt, Rose Hendrix, Suveen Ellawela, Lucas Ngoo, Joyce Chai, Zhongzheng Ren§, Ali Farhadi§, Dieter Fox§, Ranjay Krishna§
(∗ denotes equal contribution, § denotes core contributors)
Webpage •
Paper •
Code •
Models •
Data
Tether: Autonomous Functional Play with Correspondence-Driven Trajectory Warping
William Liang, Sam Wang, Hung-Ju Wang, Osbert Bastani, Yecheng Jason Ma, Dinesh Jayaraman
International Conference on Learning Representations (ICLR), 2026
Webpage •
Paper •
Code
Eurekaverse: Environment Curriculum Generation via Large Language Models
William Liang, Sam Wang, Hung-Ju Wang, Osbert Bastani, Dinesh Jayaraman†, Yecheng Jason Ma†
Conference on Robot Learning (CoRL), 2024 (Oral)
Princeton Symposium on Safe Deployment of Foundation Models in Robotics, 2024 (Poster)
Webpage •
Paper •
Code
DrEureka: Language Model Guided Sim-To-Real Transfer
Yecheng Jason Ma*, William Liang*, Hung-Ju Wang, Sam Wang, Yuke Zhu, Linxi "Jim" Fan, Osbert Bastani, Dinesh Jayaraman
Robotics: Science and Systems (RSS), 2024 (Oral)
Webpage •
Paper •
Code
Deep Learning for Rapid Cardiac Screening
Sam Wang
Regneron International Science and Engineering Fair, 2023
★ Second Grand Prize, ISEF 2023 Robotics and Intelligent Machines Category ★
★ Special Awards from Association for Computing Machinery (ACM) and International Council on Systems Engineering (INCOSE) at ISEF 2023★
★ 1st Place in Computer Science at Delaware Valley Science Fair (DVSF) ★
★ Merck Special Award for Excellence in Drug Discovery and Human Health ★
★ Janssen Cardiovascular and Metabolic Research Award ★
Webpage • Paper
PennCloud
PennCloud mimics the basic features of Google Drive and Gmail, allowing users to upload and download files, as well as send and receive mail from within the PennCloud ecosystem and externally as well. Implemented a distributed KV store Google's BigTable. The system replicates data across both a primary and secondary nodes, allowing for multinode crash fault tolerance. Each node periodically checkpoints its data to memory and logs every operation to allow for failure recovery. We also implement a multiplayer Blackjack minigame on top of our distributed backend storage and succesfully deployed PennCloud onto AWS.
Report
Machine Learning for Place Detection
Trained models to predict the location of images from only visual features. Experimented with Vision Transformers, in-context learning with vision language models, and segmentation foundation models. Won 2nd place in 200+ student competition.
Report • Models
Pennstagram Social Media Platform
Built full-stack social media application with profiles, posts, infinite feed, comments, and secure login. Implemented real-time chat using Socket.io, LLM-augmented interest search powered by GPT, YouTube-inspired recommendation algorithm with Apache Spark, and image similarity search with ChromaDB. Deployed to AWS with EC2, Livy, RDS, and S3; utilized Kafka for federated post retrieval.
Code • Report
Digital Attendance Kiosk
Built kiosk with Raspberry Pi and badge scanner to digitize attendance records for 650+ students and faculty. Developed an object-oriented interface for automated attendance tracking using Python, Tkinter, and gspread.
Code
Website Management for Local Organizations
Created and maintained brand new websites for School Newspaper, Boy Scout Troop, and community organizations. Created custom plugins for WordPress and ticket sales using PHP.
Code
Privacy Preserving Contact Tracing with Bluetooth Signals
Developed and tested a Bluetooth Low Energy (BLE)-based system for automated, privacy-preserving COVID-19 contact tracing using Raspberry Pis. Designed experiments to evaluate how factors such as distance, obstacles, and device orientation affect the Received Signal Strength Indicator (RSSI) of Bluetooth signals. Integrated additional data from gyroscope and accelerometer sensors to improve proximity detection accuracy. Built machine learning models using scikit-learn, comparing linear regression with classifiers like Decision Trees, Random Forests, and Naïve Bayes. Results demonstrated that classifiers significantly outperformed linear regression models.
Code • Report2026
CIS 5050 Distributed Systems2025
CS336 Language Modeling from Scratch (Stanford University)2024
CIS 5800 3D Computer Vision2023
CS285 Deep Reinforcement Learning (UC Berkeley)2022
CS231n Deep Learning for Computer Vision (Stanford University)I am a member of the GRASP Student, Faculty, and Industry (SFI) Seminar committee, hosting guests to give talks at our weekly seminar. So far, I've had the pleasure of hosting Chuning Zhu and Dr. Ji Woong (Brian) Kim.
I served as the Vice President of the Penn Engineering Deans' Advisory Board. Together, we organized the inaugural Penn AI conference, which brought together several hundred student and faculty attendees to explore advancements in AI, and wrote a guide for undegraduates interested in research. I'm also leading efforts to improve Penn Engineering's orientation programs.
As the Technology Chair for Penn Engineering Council, I've organized the Penn Ignition pitch competition as well as multiple seminars and panels on founding startups.
I'm an Eagle Scout from Troop 542 in Maple Glen, PA and a two time President’s Volunteer Service Award Gold recipient.
In my free time, I enjoy reading about history and philosophy, cooking, playing basketball, running, and rooting for the Minnesota Vikings.