Search papers, labs, and topics across Lattice.
3
0
5
A novel loss function that boosts convergence speed by up to 40% while improving image quality by addressing the spectral imbalance in pixel-space training.
Generative models can now handle reconstruction tasks with scarce data and traceable uncertainty by representing vector fields as neural implicit functions.
Attention's quadratic scaling problem? Solved: this new Polynomial Mixer (PoM) matches attention performance at linear complexity across diverse tasks.
