Does every sample get the same demonstrations through vote_k retriver?

[joyce0105-ops]

def vote_k_search(self):
    vote_k_idxs = self.votek_select(embeddings=self.embed_list, select_num=self.ice_num, k=self.votek_k,
                                    overlap_threshold=1)
    return [vote_k_idxs[:] for _ in range(len(self.test_ds))]

why？

[anushkasw]

I have the same question

[mihara-bot]

According to my understanding (Alg 1, Page 21, from the paper https://openreview.net/forum?id=qY1hlv7gwg), the current version only implements the step 1, which construts the candidate set for further selection. It should be followed by another TopK retriever to select from candidate set for each test input instead of being utilized for the whole test set.