Saved in:
Bibliographic Details
Main Authors: Yao, Bingsheng, Chen, Guiming, Zou, Ruishi, Lu, Yuxuan, Li, Jiachen, Zhang, Shao, Sang, Yisi, Liu, Sijia, Hendler, James, Wang, Dakuo
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2311.09782
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911822603354112
author Yao, Bingsheng
Chen, Guiming
Zou, Ruishi
Lu, Yuxuan
Li, Jiachen
Zhang, Shao
Sang, Yisi
Liu, Sijia
Hendler, James
Wang, Dakuo
author_facet Yao, Bingsheng
Chen, Guiming
Zou, Ruishi
Lu, Yuxuan
Li, Jiachen
Zhang, Shao
Sang, Yisi
Liu, Sijia
Hendler, James
Wang, Dakuo
contents While most existing works on LLM prompting techniques focus only on how to select a better set of data samples inside one single prompt input (In-Context Learning or ICL), why can not we design and leverage multiple prompts together to further improve the LLM's performance? In this work, we propose In-Context Sampling (ICS), a low-resource LLM prompting technique to produce confident predictions by optimizing the construction of multiple ICL prompt inputs. Extensive experiments with three open-source LLMs (FlanT5-XL, Mistral-7B, and Mixtral-8x7B) on four NLI datasets (e-SNLI, Multi-NLI, ANLI, and Contract-NLI) and one QA dataset (CommonsenseQA) illustrate that ICS can consistently enhance LLMs' performance. An in-depth evaluation with three data similarity-based ICS strategies suggests that these strategies can further elevate LLM's performance, which sheds light on a new yet promising future research direction.
format Preprint
id arxiv_https___arxiv_org_abs_2311_09782
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle More Samples or More Prompts? Exploring Effective In-Context Sampling for LLM Few-Shot Prompt Engineering
Yao, Bingsheng
Chen, Guiming
Zou, Ruishi
Lu, Yuxuan
Li, Jiachen
Zhang, Shao
Sang, Yisi
Liu, Sijia
Hendler, James
Wang, Dakuo
Computation and Language
While most existing works on LLM prompting techniques focus only on how to select a better set of data samples inside one single prompt input (In-Context Learning or ICL), why can not we design and leverage multiple prompts together to further improve the LLM's performance? In this work, we propose In-Context Sampling (ICS), a low-resource LLM prompting technique to produce confident predictions by optimizing the construction of multiple ICL prompt inputs. Extensive experiments with three open-source LLMs (FlanT5-XL, Mistral-7B, and Mixtral-8x7B) on four NLI datasets (e-SNLI, Multi-NLI, ANLI, and Contract-NLI) and one QA dataset (CommonsenseQA) illustrate that ICS can consistently enhance LLMs' performance. An in-depth evaluation with three data similarity-based ICS strategies suggests that these strategies can further elevate LLM's performance, which sheds light on a new yet promising future research direction.
title More Samples or More Prompts? Exploring Effective In-Context Sampling for LLM Few-Shot Prompt Engineering
topic Computation and Language
url https://arxiv.org/abs/2311.09782