I saw this paper’s abstract and was reminded of your comment: https://arxiv.org/abs/2212.10559 https://arxiv.org/abs/2212.10559 > Experimentally, we comprehens

I saw this paper’s abstract and was reminded of your comment: https://arxiv.org/abs/2212.10559 https://arxiv.org/abs/2212.10559 > Experimentally, we comprehensively compare the behavior of ICL and explicit finetuning based on real tasks to provide empirical evidence that supports our understanding