12 papers
Grounding Computer Use Agents on Human Demonstrations
Aarash Feizi, Shravan Nayak, Xiangru Jian +14
Building reliable computer-use agents requires grounding: accurately connecting natural language instructions to the correct on-screen elements. While large datasets exist for web…
Do Enterprise Systems Need Learned World Models? The Importance of Context to Infer Dynamics
Jishnu Sethumadhavan Nair, Patrice Bechard, Rishabh Maheshwary +14
World models enable agents to anticipate the effects of their actions by internalizing environment dynamics. In enterprise systems, however, these dynamics are often defined by ten…
Multi-scale Predictive Representations for Goal-conditioned Reinforcement Learning
Valliappan Chidambaram Adaikkappan, David Meger, Sai Rajeswar +1
This paper investigates robust representation learning in offline goal-conditioned reinforcement learning (GCRL). Particularly in sparse reward scenarios, learning representations…
VectorGym: A Multitask Benchmark for SVG Code Generation, Sketching, and Editing
Juan Rodriguez, Haotian Zhang, Abhay Puri +13
We introduce VectorGym, a comprehensive benchmark suite for Scalable Vector Graphics (SVG) that spans generation from text and sketches, complex editing, and visual understanding.…
StarFlow: Generating Structured Workflow Outputs From Sketch Images
Patrice Bechard, Chao Wang, Amirhossein Abaskohi +6
Workflows are a fundamental component of automation in enterprise platforms, enabling the orchestration of tasks, data processing, and system integrations. Despite being widely use…
Rendering-Aware Reinforcement Learning for Vector Graphics Generation
Juan A. Rodriguez, Haotian Zhang, Abhay Puri +12
Scalable Vector Graphics (SVG) offer a powerful format for representing visual designs as interpretable code. Recent advances in vision-language models (VLMs) have enabled high-qua…