Lune

EMNLP Findings2023

Meta-Learning of Prompt Generation for Lightweight Prompt Engineering on Language-Model-as-a-Service

Hyeonmin Ha, Jihye Lee, Wookje Han, Byung-Gon Chun

摘要

Recently, many companies have been providing the capabilities of large language models as services. These Language-Model-as-a-Service (LMaaS) offerings support a variety of user tasks through in-context learning from prompts, which include instructions and demonstrations of the task. However, for users, manually crafting prompts or running automatic prompt tuning methods themselves can be demanding. Despite these challenges, LMaaS providers do not offer automatic prompt engineering methods as part of their services. One of the major obstacles to deploying them on an LMaaS is the heavy computational costs associated with automatic prompt engineering methods. These methods are typically designed to iterate through tens of thousands of examples, which impose unaffordable overheads for LMaaS providers. In this paper, we introduce MetaL-Prompt, a novel lightweight automatic prompt generation method for LMaaS. MetaL-Prompt meta-trains a prompt generation model (PGM) to enable robust learning by the language model from the contexts created by the generated prompts (i.e., in-context learning). Thanks to our meta-learning approach, a PGM can generate prompts for unseen tasks without requiring additional training for those specific tasks. Furthermore, the PGM can generate prompts with a single forward pass, significantly reducing computational costs compared to previous methods. We evaluate MetaL-Prompt on a range of unseen tasks and find that it improves performance by up to 19.4% in terms of mean F1 score on QA datasets compared to the state-of-the-art baseline P-tuning, with limited computational cost.

问 Lune

几分钟读透这篇 Paper,不用再花一下午。

Lune 能读完全文,也能读引用它的顶会 Paper,讲清它的结论、证据和局限,每一点都注明出自哪篇 Paper。不用硬啃,也不靠猜。

已有 1 篇顶会 Paper 引用了它。别等审稿人问起,先弄清它们发现了什么。

可以从这些问题问起

Lune 可以查阅

全文 · 引用它的顶会 Paper · 相关的 Paper

获取答案

免费开始,无需绑卡

脉络

  • Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity 引用了 Calibrate Before Use: Improving Few-shot Performance of Language Models
它引用的引用它的从左到右按年份排列,引用越多,位置越高。

引用它的顶会 Paper1

  1. CAP: Controllable Alignment Prompting for Unlearning in LLMsZhaokun Wang, Jinyu Guo, Jingwen Pu, Hongli Pu 等ACL 2026

它引用的 Paper16

  1. LoRA: Low-Rank Adaptation of Large Language ModelsEdward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu 等ICLR 2022 · 18,833 次引用
  2. Calibrate Before Use: Improving Few-shot Performance of Language ModelsZihao Zhao, Eric Wallace, Shi Feng, Dan Klein 等ICML 2021 · 1,843 次引用
  3. Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order SensitivityYao Lu, Max Bartolo, Alastair Moore, Sebastian Riedel 等ACL 2022 · 1,494 次引用
  4. AutoPrompt: Eliciting Knowledge from Language Models with Automatically Generated PromptsTaylor Shin, Yasaman Razeghi, Robert L. Logan IV, Eric Wallace 等EMNLP 2020 · 1,162 次引用
  5. Orca: A Distributed Serving System for Transformer-Based Generative ModelsGyeong-In Yu, Joo Seong Jeong, Geon-Woo Kim, Soojeong Kim 等OSDI 2022 · 690 次引用

相关 Paper

  1. Black-Box Tuning for Language-Model-as-a-ServiceTianxiang Sun, Yunfan Shao, Hong Qian, Xuanjing Huang 等ICML 2022 · 343 次引用
  2. SEE: Strategic Exploration and Exploitation for Cohesive In-Context Prompt OptimizationWendi Cui, Jiaxin Zhang, Zhuohang Li, Hao Sun 等ACL 2025
  3. λ-Tune: Harnessing Large Language Models for Automated Database System TuningVictor Giannakouris, Immanuel TrummerSIGMOD 2025 · 20 次引用
  4. Prompt-MII: Meta-Learning Instruction Induction for LLMsEmily Xiao, Yixiao Zeng, Ada Chen, Chin-Jou Li 等ICLR 2026 · 9 次引用
  5. Prompt Optimization with EASE? Efficient Ordering-aware Automated Selection of ExemplarsZhaoxuan Wu, Xiaoqiang Lin, Zhongxiang Dai, Wenyang Hu 等NeurIPS 2024 · 44 次引用
黄昏的海面,两侧是细线勾勒的悬崖