Di Wang

Evolving Cache Schedules can slash action-generation time by over 8x without sacrificing performance, revolutionizing real-time deployment of diffusion policies.

Preserving full-band spectral information in hyperspectral image classification can significantly boost model performance while keeping parameter tuning minimal.

Latent reasoning methods may seem unfaithful at convergence, but their causal contributions to answers decay over training, revealing hidden complexities in model behavior.