Yes but how did you create the embeddings ? Would have loved a deeper understanding of concepts, like which kind of model to use for such use cases, what about large texts, etc
Is there a research comparing the Similarity between two sentences based on Warping and levenshtein distance on a word by word embeddings versus the embedding of the whole sentence ?
I think it’s not about the exact words being similar in a sentence, but the semantic meaning of the sentences. The fact that we are looking at multiple-dimensional space here, these sentences indeed would come out to be similar
Let's say one two movies have value for Action and Comedy 0, 1 the dot product will be 1. lets say another two movies have value of 0, 2 for both movies. And it will get dot product of 4. And will conclude second set is more simmilor than first set. But it is not the case in real. Could you please explain this.
Luis, thanks for making the important concept of embeddings and similarity simple and intuitive
Yes but how did you create the embeddings ?
Would have loved a deeper understanding of concepts, like which kind of model to use for such use cases, what about large texts, etc
Is there a research comparing the Similarity between two sentences based on Warping and levenshtein distance on a word by word embeddings versus the embedding of the whole sentence ?
great content. I would be curious as to how Cohere computes their embeddings. How can we know if embeddings are good.
Question. What about two completely different sentences that say the same thing but with completely different words?
I think it’s not about the exact words being similar in a sentence, but the semantic meaning of the sentences. The fact that we are looking at multiple-dimensional space here, these sentences indeed would come out to be similar
Let's say one two movies have value for Action and Comedy 0, 1 the dot product will be 1. lets say another two movies have value of 0, 2 for both movies. And it will get dot product of 4. And will conclude second set is more simmilor than first set. But it is not the case in real. Could you please explain this.