We-Math: Does Your Large Multimodal Model Achieve Human-like Mathematical Reasoning? Paper • 2407.01284 • Published Jul 1 • 75 • 9