Self-ICL: Zero-Shot In-Context Learning with Self-Generated Demonstrations
This addresses the challenge for end-users who query language models without demonstration access, offering a zero-shot solution that is incremental but practical.
The paper tackles the problem of in-context learning (ICL) for large language models without access to demonstration pools by introducing Self-ICL, a framework that generates pseudo-demonstrations from the model itself, resulting in improved performance over zero-shot baselines on 23 BIG-Bench Hard tasks and achieving results comparable to using real demonstrations with zero-shot chain-of-thought.
Large 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.