さらなる注釈データを必要とするため困難 2 導入:テキスト埋め込み [1]Muennighoff et al.:Massive Text Embedding Benchmark (EACL’23) [2]Thakur et al.;BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information Retrieval Models(NeurlPS’21)
et al.:Multitask Prompted Training Enables Zero-Shot Task Generalization(arXiv’22) [6]Zhou et al.: Prompt Consistency for Zero-Shot Task Generalization(arXiv’22)