Chen, Wei LinWei LinChenWu, Cheng KuangCheng KuangWuYUN-NUNG CHENHSIN-HSI CHEN2024-03-072024-03-072023-01-019798891760608https://scholars.lib.ntu.edu.tw/handle/123456789/640564https://www.scopus.com/inward/record.uri?eid=2-s2.0-85184817497&doi=10.18653%2fv1%2f2023.emnlp-main.968&partnerID=40&md5=08b0b8d8ef943e89dc6fe8a559d96b86Large language models (LLMs) have exhibited striking in-context learning (ICL) ability to adapt to target tasks with a few input-output demonstrations. For better ICL, different methods are proposed to select representative demonstrations from existing training corpora. However, such settings are not aligned with real-world practices, as end-users usually query LMs without access to demonstration pools. In this work, we introduce SELF-ICL-a simple framework which bootstraps LMs' intrinsic capabilities to perform zero-shot ICL. Given a test input, SELF-ICL first prompts the model to generate pseudo-inputs. Next, the model predicts pseudo-labels for the pseudo-inputs via zero-shot prompting. Finally, we perform ICL for the test input with the pseudo-input-label pairs as demonstrations. Evaluation on 23 BIG-Bench Hard tasks shows SELF-ICL outperforms zero-shot baselines on both average accuracy and head-to-head comparison. Moreover, with zero-shot chain-of-thought, SELF-ICL achieves results comparable to using real demonstrations. Additionally, we conduct a range of analyses to validate SELF-ICL's effectiveness and provide insights for its behaviors under different settings.conference paper2-s2.0-85184817497https://api.elsevier.com/content/abstract/scopus_id/85184817497