Róbert Belanec

Paraphrases of the max 500 tokens subset of the MMLU dataset. We train models on both paraphrases and not paraphrases to increase robustness.

Collection of all soft-prompts used in the paper https://arxiv.org/abs/2408.01119