用已有任务示例指导新任务，无需人工标注。
CrossICL: Cross-Task In-Context Learning via Unsupervised Demonstration Transfer
- 通过两阶段对齐策略减少任务间差异干扰。
- 在875个NLP任务上验证有效，提升模型泛化能力。
- 适合缺乏标注数据的场景，尤其对大模型应用者有价值。
上下文学习（ICL）通过示范样例提升大语言模型性能，但示范获取依赖大量人工投入。在多数实际场景中，用户难以或不愿提供示范。受人类类比思维启发，本文提出新型ICL范式CrossICL，探索如何复用已有源任务示范来指导目标任务，实现无需额外人工标注的可靠引导。为此，我们设计两阶段对齐策略，以缓解任务间差异带来的干扰，作为实验基础。在此基础上，基于Super-NI基准的875个NLP任务及六类大语言模型（包括GPT-4o）开展系统性研究。实验表明CrossICL有效，并揭示了跨任务示范选择标准及任务差距引发的干扰类型等关键洞见。
原文摘要 · Abstract (English)
In-Context Learning (ICL) enhances the performance of large language models (LLMs) with demonstrations. However, obtaining these demonstrations primarily relies on manual effort. In most real-world scenarios, users are often unwilling or unable to provide such demonstrations. Inspired by the human analogy, we explore a new ICL paradigm CrossICL to study how to utilize existing source task demonstrations in the ICL for target tasks, thereby obtaining reliable guidance without any additional manual effort. To explore this, we first design a two-stage alignment strategy to mitigate the interference caused by gaps across tasks, as the foundation for our experimental exploration. Based on it, we conduct comprehensive exploration of CrossICL, with 875 NLP tasks from the Super-NI benchmark and six types of LLMs, including GPT-4o. Experimental results demonstrate the effectiveness of CrossICL and provide valuable insights on questions like the criteria for selecting cross-task demonstrations, as well as the types of task-gap-induced interference in CrossICL.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。