Robert Sim

Learning GUI Grounding with Spatial Reasoning from Visual Feedback

PE-SGD: Differentially Private Deep Learning via Evolution of Gradient Subspace for Text

ACON: Optimizing Context Compression for Long-horizon LLM Agents

Simulating Environments with Large Language Models for Generic Agent Training

Contextual Integrity in LLMs via Reasoning and Reinforcement Learning

Learning to Specialize: Joint Gating-Expert Training for Adaptive MoEs in Decentralized Settings

VLMGuard: Defending VLMs against Malicious Prompts via Unlabeled Data

Exploring How LLMs Capture and Represent Domain-Specific Knowledge