提出统一度量标准，提升上下文学习中示范样本选择效果
Affinity and Diversity: A Unified Metric for Demonstration Selection via Internal Representations
- 基于模型内部表征，同时衡量示范样本的亲和性与多样性
- 亲和性与多样性均与测试准确率强相关，验证有效性
- 可统一现有不同方法，解决示范选择结果不一致问题
上下文学习（ICL）性能高度依赖示范样本的选择。现有示范选择方法优化目标各异，导致结果不一致。为此，我们提出一种统一度量——亲和性与多样性，该度量利用ICL模型的内部表示。实验表明，亲和性与多样性均与测试准确率显著相关，证明其在示范选择中的有效性。此外，我们的度量与多种先前工作高度一致，有效统一了不同方法间的分歧。
原文摘要 · Abstract (English)
The performance of In-Context Learning (ICL) is highly sensitive to the selected demonstrations. Existing approaches to demonstration selection optimize different objectives, yielding inconsistent results. To address this, we propose a unified metric--affinity and diversity--that leverages ICL model's internal representations. Our experiments show that both affinity and diversity strongly correlate with test accuracies, indicating their effectiveness for demonstration selection. Moreover, we show that our proposed metrics align well with various previous works to unify the inconsistency.
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。