4 ms·
Don’t the high end embedding services use a transformer with attention to compute embeddings? If so, I thought that would indeed capture the semantic meaning qu
by manmal 1y ago
Don’t the high end embedding services use a transformer with attention to compute embeddings? If so, I thought that would indeed capture the semantic meaning quite well, including the trait-is-described-by-direction-vector.
- realbenpope 1y agoYou are correct. https://deepmind.google/research/publications/157741/ https://deepmind.google/research/publications/157741/