PhD student, Tsinghua University, Tsinghua University
1 paper at NeurIPS 2025
We propose a more expressive Transformer, which exceeding its original $TC^0$ expressiveness.