@inproceedings{liang-etal-2024-self,
title = "Self-Renewal Prompt Optimizing with Implicit Reasoning",
author = "Liang, Zihan and
Chen, Ben and
Ran, Zhuoran and
Wang, Zihan and
Dai, Huangyu and
Ma, Yufei and
Gao, Dehong and
Cai, Xiaoyan and
Yang, Libin",
editor = "Al-Onaizan, Yaser and
Bansal, Mohit and
Chen, Yun-Nung",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2024",
month = nov,
year = "2024",
address = "Miami, Florida, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.findings-emnlp.171",
pages = "3030--3041",
abstract = "The effectiveness of Large Language Models (LLMs) relies on their capacity to understand instructions and generate human-like responses. However, aligning LLMs with complex human preferences remains a significant challenge due to the potential misinterpretation of user prompts. Current methods for aligning LLM behaviors fall into two categories: output optimization (such as RLHF, RLAIF, and DPO) and input optimization (like OPRO and BPO). While both approaches aim to guide LLMs towards generating responses that align with desired objectives, the labor-intensive and intentions-inconsistent data annotation, as well as the strict and tedious training supervision, make them struggle to yield optimal results across all models. To address these shortcomings, we introduce a novel self-renewal approach called Prompt Optimization with Implicit Reasoning (POIR). It consists of two key components: 1) a model-specific and self-recirculating data collection method that leverages self-evaluation to enhance prompts in accordance with the model{'}s intrinsic logits, and 2) a prompt rewrite schema that injects implicit reasoning for direct preference learning. Through self-renewal optimization, POIR refines LLM outputs to better align with human preferences across various LLMs and tasks, without relying on supervised fine-tuning. Extensive experiments on a range of LLMs and tasks demonstrate POIR{'}s superior performance. We believe this advancement offers a novel paradigm for developing LLMs that are more attuned to user intentions.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="liang-etal-2024-self">
<titleInfo>
<title>Self-Renewal Prompt Optimizing with Implicit Reasoning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Zihan</namePart>
<namePart type="family">Liang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ben</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhuoran</namePart>
<namePart type="family">Ran</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zihan</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Huangyu</namePart>
<namePart type="family">Dai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yufei</namePart>
<namePart type="family">Ma</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dehong</namePart>
<namePart type="family">Gao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiaoyan</namePart>
<namePart type="family">Cai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Libin</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2024</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yaser</namePart>
<namePart type="family">Al-Onaizan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohit</namePart>
<namePart type="family">Bansal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yun-Nung</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Miami, Florida, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>The effectiveness of Large Language Models (LLMs) relies on their capacity to understand instructions and generate human-like responses. However, aligning LLMs with complex human preferences remains a significant challenge due to the potential misinterpretation of user prompts. Current methods for aligning LLM behaviors fall into two categories: output optimization (such as RLHF, RLAIF, and DPO) and input optimization (like OPRO and BPO). While both approaches aim to guide LLMs towards generating responses that align with desired objectives, the labor-intensive and intentions-inconsistent data annotation, as well as the strict and tedious training supervision, make them struggle to yield optimal results across all models. To address these shortcomings, we introduce a novel self-renewal approach called Prompt Optimization with Implicit Reasoning (POIR). It consists of two key components: 1) a model-specific and self-recirculating data collection method that leverages self-evaluation to enhance prompts in accordance with the model’s intrinsic logits, and 2) a prompt rewrite schema that injects implicit reasoning for direct preference learning. Through self-renewal optimization, POIR refines LLM outputs to better align with human preferences across various LLMs and tasks, without relying on supervised fine-tuning. Extensive experiments on a range of LLMs and tasks demonstrate POIR’s superior performance. We believe this advancement offers a novel paradigm for developing LLMs that are more attuned to user intentions.</abstract>
<identifier type="citekey">liang-etal-2024-self</identifier>
<location>
<url>https://aclanthology.org/2024.findings-emnlp.171</url>
</location>
<part>
<date>2024-11</date>
<extent unit="page">
<start>3030</start>
<end>3041</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Self-Renewal Prompt Optimizing with Implicit Reasoning
%A Liang, Zihan
%A Chen, Ben
%A Ran, Zhuoran
%A Wang, Zihan
%A Dai, Huangyu
%A Ma, Yufei
%A Gao, Dehong
%A Cai, Xiaoyan
%A Yang, Libin
%Y Al-Onaizan, Yaser
%Y Bansal, Mohit
%Y Chen, Yun-Nung
%S Findings of the Association for Computational Linguistics: EMNLP 2024
%D 2024
%8 November
%I Association for Computational Linguistics
%C Miami, Florida, USA
%F liang-etal-2024-self
%X The effectiveness of Large Language Models (LLMs) relies on their capacity to understand instructions and generate human-like responses. However, aligning LLMs with complex human preferences remains a significant challenge due to the potential misinterpretation of user prompts. Current methods for aligning LLM behaviors fall into two categories: output optimization (such as RLHF, RLAIF, and DPO) and input optimization (like OPRO and BPO). While both approaches aim to guide LLMs towards generating responses that align with desired objectives, the labor-intensive and intentions-inconsistent data annotation, as well as the strict and tedious training supervision, make them struggle to yield optimal results across all models. To address these shortcomings, we introduce a novel self-renewal approach called Prompt Optimization with Implicit Reasoning (POIR). It consists of two key components: 1) a model-specific and self-recirculating data collection method that leverages self-evaluation to enhance prompts in accordance with the model’s intrinsic logits, and 2) a prompt rewrite schema that injects implicit reasoning for direct preference learning. Through self-renewal optimization, POIR refines LLM outputs to better align with human preferences across various LLMs and tasks, without relying on supervised fine-tuning. Extensive experiments on a range of LLMs and tasks demonstrate POIR’s superior performance. We believe this advancement offers a novel paradigm for developing LLMs that are more attuned to user intentions.
%U https://aclanthology.org/2024.findings-emnlp.171
%P 3030-3041
Markdown (Informal)
[Self-Renewal Prompt Optimizing with Implicit Reasoning](https://aclanthology.org/2024.findings-emnlp.171) (Liang et al., Findings 2024)
ACL
- Zihan Liang, Ben Chen, Zhuoran Ran, Zihan Wang, Huangyu Dai, Yufei Ma, Dehong Gao, Xiaoyan Cai, and Libin Yang. 2024. Self-Renewal Prompt Optimizing with Implicit Reasoning. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 3030–3041, Miami, Florida, USA. Association for Computational Linguistics.