Search papers, labs, and topics across Lattice.
Affiliation:
3
0
5
3
Achieving over 99% performance retention while significantly accelerating large recommendation models by intelligently merging experts based on functional similarity.
By cleverly combining linear and softmax attention, HyTRec achieves state-of-the-art recommendation accuracy on long sequences while maintaining linear inference speed, resolving a key tradeoff in the field.
Current vision-language models stumble on subtle spine diagnoses, but a new dataset and benchmark expose these weaknesses and pave the way for clinically useful AI.
