Syn-GRPO: Self-Evolving Data Synthesis for MLLM Perception Reasoning

Re-assign community