Depth-Breadth Synergy in RLVR: Unlocking LLM Reasoning Gains with Adaptive Exploration Paper • 2508.13755 • Published Aug 19, 2025 • 14