A user calls that prospect a tragedy, comparing it with generative text models’ focus on conversational rollout.
A user worries that embedding models could become narrowly optimized for retrieval, just as they say generative text models have become narrowly optimized for conversational rollout. They argue there is much more to explore in representations.
—
2 Sources, first seen ago
@thesephistIt would be a tragedy for embedding models to become narrowly optimized for retrieval in the same way that generative text models have become narrowly optimized for conversational rollout. There is much magic to explore in the realm of representations.
@charles_irlRT @thesephist: It would be a tragedy for embedding models to become narrowly optimized for retrieval in the same way that generative text…@thesephistIt would be a tragedy for embedding models to become narrowly optimized for retrieval in the same way that generative text models have become narrowly optimized for conversational rollout. There is much magic to explore in the realm of representations.@charles_irlRT @thesephist: It would be a tragedy for embedding models to become narrowly optimized for retrieval in the same way that generative text…@thesephistIt would be a tragedy for embedding models to become narrowly optimized for retrieval in the same way that generative text models have become narrowly optimized for conversational rollout. There is much magic to explore in the realm of representations.@charles_irlRT @thesephist: It would be a tragedy for embedding models to become narrowly optimized for retrieval in the same way that generative text…@thesephistIt would be a tragedy for embedding models to become narrowly optimized for retrieval in the same way that generative text models have become narrowly optimized for conversational rollout. There is much magic to explore in the realm of representations.@charles_irlRT @thesephist: It would be a tragedy for embedding models to become narrowly optimized for retrieval in the same way that generative text…—
Not ranked yet
—
Not ranked yet
