Tho Quan

Causal memory can boost execution accuracy in Text-to-SQL tasks, but its effectiveness is context-dependent and not universally superior to other retrieval methods.

Lower-order discontinuities in Sparse Mixture-of-Experts architectures dominate input space, significantly affecting model outputs and performance.

Transformers can nail in-context learning for regression even when the data is a mess of non-Gaussian noise, heavy tails, and non-i.i.d. distributions, outperforming classical estimators.