The Last Mile to the Real World: Agent-Native Primitives as Reusable Building Blocks for Solving Real-World Tasks Paper • 2611.07461 • Published 4 days ago • 25
PACEvolve++: Improving Test-time Learning for Evolutionary Search Agents Paper • 2605.07039 • Published May 7 • 6
TUA-Bench: A Benchmark for General-Purpose Terminal-Use Agents Paper • 2606.28480 • Published Jun 26 • 49