Search papers, labs, and topics across Lattice.
Affiliation:
2
0
5
2
Recurrent token mixing via GTR achieves 58.9 AP on COCO with sub-2ms latency, offering a practical, efficient alternative to quadratic self-attention for high-resolution dense prediction.
Post-norm normalization significantly enhances performance in LLMs when depth is introduced through a curriculum, outperforming pre-norm by an order of magnitude.
