Bo Xu

This paper develops a connectome-informed neural network algorithm that improves RNN and LLM performance by incorporating natural cortical connectivity motifs.

The paper investigates dioicy-monoicy transitions in liverworts, confirming monoicy evolved from a dioicous ancestor and describing potential re-evolution of dioicy with new sex chromosome formation.

McAN constructs haplotype networks using a minimum-cost arborescence approach, achieving speeds two orders of magnitude faster than existing methods.