6 papers
RHO: Your Coding Agent is Secretly a Roboticist
Karim Elmaaroufi, Justin Svegliato, Sarunas Kalade +3
Code-as-Policies (CaP) has shown that large language models (LLMs) can write code to solve robotics tasks by composing perception, planning, and control primitives. Recent CaP syst…
Active teacher selection for reward learning
Rachel Freedman, Justin Svegliato, Kyle Wray +1
Reward learning techniques enable machine learning systems to learn objectives from human feedback. A core limitation of these systems is their assumption that all feedback comes f…
Fine-Tuning LLMs with Fine-Grained Human Feedback on Text Spans
Sky CH-Wang, Justin Svegliato, Helen Appel +1
We present a method and dataset for fine-tuning language models with preference supervision using feedback-driven improvement chains. Given a model response, an annotator provides…
GRAID: Enhancing Spatial Reasoning of VLMs Through High-Fidelity Data Generation
Karim Elmaaroufi, Liheng Lai, Justin Svegliato +3
Vision Language Models (VLMs) achieve strong performance on many vision-language tasks but often struggle with spatial reasoning$\unicode{x2014}$a prerequisite for many application…
AssistanceZero: Scalably Solving Assistance Games
Cassidy Laidlaw, Eli Bronstein, Timothy Guo +5
Assistance games are a promising alternative to reinforcement learning from human feedback (RLHF) for training AI assistants. Assistance games resolve key drawbacks of RLHF, such a…
MICE for CATs: Model-Internal Confidence Estimation for Calibrating Agents with Tools
Nishant Subramani, Jason Eisner, Justin Svegliato +3
Tool-using agents that act in the world need to be both useful and safe. Well-calibrated model confidences can be used to weigh the risk versus reward of potential actions, but pri…