PhD student, Nanjing University
2 papers at NeurIPS 2025
We identify the output projection module as the key component enabling reasoning in LLMs, and propose Stethoscope for Networks (SfN) to diagnose and support this claim.
We systematically investigate the design space and scaling property of native Multimodal Large Language Models and introduce a novel MLLM that achieves competitive performance against existing MLLMs.
