[论文] EEVEE: Towards Test-time Prompt Learning in the Real World for Self-Im...

论文概要

研究领域: ML
作者: Weixian Xu, Shilong Liu, Mengdi Wang
发布时间: 2026-06-09
arXiv: 2606.11182

中文摘要

EEVEE是首个面向LLM智能体的多数据集测试时提示学习框架，支持真实世界任务流中的测试时提示学习。引入路由器将输入划分为任务簇并分配给合适的提示配置，通过路由器-提示协同进化策略优化。实验表明，相比Qwen3-4B-Instruct和DeepSeek-V3.2，平均多基准分数分别提升10.38和24.32分，超越SOTA方法GEPA和ACE达37.2%和48.2%。

原文摘要

In this paper, we propose EEVEE, the first multi-dataset test-time prompt learning framework for LLM agents, enabling test-time prompt learning under real-world task streams. Existing methods are largely designed for single-dataset settings, while real-world applications require models to handle heterogeneous input streams drawn from multiple datasets, domains, and task distributions, limiting their practical applicability. To mitigate cross-dataset interference, EEVEE introduces a router that partitions incoming inputs into task簇 and assigns them to suitable prompt configurations. This design is optimized via a router-prompt co-evolution strategy, which employs interleaved router and prompt learning phases to address their mutual dependency. Experiments across multiple datasets demonstrat...

自动采集于 2026-06-11

#论文 #arXiv #ML #小凯