collaborators

40 papers

cs.LG2026

Q-Learning With World Models

Perry Dong, Yueru Jia, Chelsea Finn +1

Off-policy reinforcement learning (RL) has become increasingly sample-efficient, enabling applications such as RL fine-tuning of Vision-Language-Action models into reliable, high-p…

cs.RO2026

Cross-Embodiment Transfer via Behavior-Aligned Representations

Ajay Sridhar, Jensen Gao, Jonathan Yang +3

The paper investigates how behavior-aligned representations such as object bounding boxes, language-described motions, and end-effector traces can improve cross-embodiment transfer…

cs.LG2026

Do You Really Need to Pretrain Q-Functions for Online RL Fine-Tuning?

Perry Dong, Ron Polonsky, Dorsa Sadigh +2

The paper investigates whether pretraining Q-functions is beneficial when fine‑tuning a pretrained policy in online reinforcement learning, finding that naive Q‑function pretrainin…

cs.RO2026

VIA: Visual Interface Agent for Robot Control

Hengyuan Hu, Priya Sundaresan, Jensen Gao +1

The paper introduces VIA, a framework that lets a pre‑trained foundation model control a robot by interacting with a browser‑based 3D visual interface, achieving high zero‑shot suc…

cs.AI2026

SPIRAL: Learning to Search and Aggregate

Jubayer Ibn Hamid, Ifdita Hasan Orney, Michael Y. Li +5

Language model reasoning can be substantially improved at test time via scaffolds that scale inference compute across different primitives -- sequential reasoning within a trace, i…

cs.RO2026

RoboCade: Gamifying Robot Data Collection

Suvir Mirchandani, Mia Tang, Jiafei Duan +3

Imitation learning from human demonstrations has become a dominant approach for training autonomous robot policies. However, collecting demonstration datasets is costly: it often r…