Reduce Retraining by Recycling Prompts - Analytics Vidhya

Notably, both methods are based on correspondences between the embedding representations of tokens across the two models. It was hypothesized that a recycler trained to map embeddings from Ms to Mt can also be used to map prompts.