Search papers, labs, and topics across Lattice.
1
0
2
Ditch one-hot encoding: LLM embeddings give your AI a deeper understanding of building semantics, boosting F1-scores by nearly 3% when classifying building object subtypes.
