Search papers, labs, and topics across Lattice.
Renmin University of China
3
0
6
Decoupling statistical preference learning from intent parsing allows personal agents to adapt to user preferences with unprecedented efficiency and accuracy.
Adjusting gate sharpness based on routing confidence leads to significant performance gains in Mixture-of-Experts models without the burden of extra parameters.
Achieve high-fidelity video generation without compromising reasoning by progressively handing off generation from a lightweight, reasoning-aligned generator to a high-capacity pretrained generator in a shared latent space.
