Scaling Agentic Capabilities, Not Context: Efficient Reinforcement Finetuning for Large Toolspaces Paper • 2603.06713 • Published 16 days ago • 16