Search papers, labs, and topics across Lattice.
1
0
2
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.
