ProCUA-SFT: Automatically Generated Training Data for Desktop Agents17. June 20264. July 2026AI ModelsAutomatically synthesized training data improves desktop agents by 18.7 percentage points compared to previous approaches. Share on:
SpatialClaw: Code-Based Interface for Spatial Reasoning in AI Agents12. June 20264. July 2026AI ModelsCode-based interfaces instead of rigid tool-calls enable AI agents to analyze spatial scenes more flexibly and solve complex 3D/4D tasks iteratively. Share on:
Astra: Spatial Reasoning Through Simulated Visual Evidence8. June 20264. July 2026AI ModelsAstra combines an RL-trained vision-language model with a world simulator to improve spatial reasoning through selectively generated perspectives. Share on:
