上下文学习的数据高效样本选择
Data Efficient Sample Selection for In-Context Learning
浏览论文内容
中文总结 AI 辅助
针对上下文学习中的示例选择，提出DearICL框架，将选择视为子集排序问题，采用可微排序和间隙索引bandit，在开源LLM基准上显著提升准确率且样本高效。
中文摘要 AI 辅助
上下文学习（ICL）范式帮助大型语言模型（LLMs）无需微调即可适应新任务。然而，从大量示例子集中选择最优的演示示例组合是一个具有挑战性的问题。现有的选择方法未能建模ICL样本与下游LLM性能之间的复杂关系。它们通常执行静态的任务级选择，即离线选择子集，这可能导致无法泛化到未见过的查询。我们提出了DearICL（数据高效排序算法）用于ICL样本，这是一个将演示示例选择建模为子集排序问题的新框架。DearICL采用了一个非线性替代模型，该模型在间隙索引（gap-index）bandit算法中使用了可微排序目标。基于间隙索引的方法能够对好的臂和边界臂进行细粒度区分，这被用作辅助目标，通过对边界臂进行充分采样来训练非线性替代模型，从而支持实例级子集排序。在使用开源LLMs的示例选择基准测试中，DearICL相较于强线性bandit基线实现了8.08%至15.9%的准确率提升，且样本复杂度较低。代码和数据：此https URL。
英文摘要
The In-context learning (ICL) paradigm aids large language models (LLMs) to adapt to new tasks without need for fine-tuning. However, selecting an optimal combination of demonstration examples from a large pool of example subsets is a challenging problem. Existing approaches for selection do not model the complex relationship between ICL samples and downstream LLM performance. They typically perform static task-level selection, choosing subsets once offline, which can fail to generalize to unseen queries. We introduce DearICL (Data Efficient Algorithm for Ranking) ICL samples, a new framework that models demonstration example selection as a subset ranking problem. DearICL employs a non-linear surrogate employing a differentiable sorting objective within a gap-index bandit algorithm. The gap-index based approach enables fine-grained separation of good arms and borderline arms, which is used as an auxiliary objective to train the non-linear surrogate through sufficient sampling of borderline arms, supporting instance-level subset ranking. On exemplar selection benchmarks with open-source LLMs, DearICL achieves 8.08-15.9% accuracy gains over strong linear bandit baselines, with low sample complexity. Code and data: https://github.com/VenkteshV/DearICL.
发表机构
- Stockholm University（斯德哥尔摩大学）
- SKIM Group B.V.（SKIM集团有限公司）
- TU Delft（代尔夫特理工大学）
机构由 AI 辅助整理，请以论文原文为准。
