Search papers, labs, and topics across Lattice.
3
0
6
Bi-CFM achieves over two orders of magnitude speedup while improving accuracy in solving inverse problems of chaotic systems, even under conservation constraints.
Simple MLPs pre-trained on low-resolution spectra can generalize surprisingly well to higher-resolution data for stellar parameter estimation, sometimes outperforming transformer-based embeddings.
Historical astronomical literature contains predictive structure for future concept-object associations that is missed by both recency heuristics and neighborhood-based methods.
