@inproceedings{lahoti-etal-2023-improving,
title = "Improving Diversity of Demographic Representation in Large Language Models via Collective-Critiques and Self-Voting",
author = "Lahoti, Preethi and
Blumm, Nicholas and
Ma, Xiao and
Kotikalapudi, Raghavendra and
Potluri, Sahitya and
Tan, Qijun and
Srinivasan, Hansa and
Packer, Ben and
Beirami, Ahmad and
Beutel, Alex and
Chen, Jilin",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.643",
doi = "10.18653/v1/2023.emnlp-main.643",
pages = "10383--10405",
abstract = "A crucial challenge for generative large language models (LLMs) is diversity: when a user{'}s prompt is under-specified, models may follow implicit assumptions while generating a response, which may result in homogenization of the responses, as well as certain demographic groups being under-represented or even erased from the generated responses. In this paper, we formalize the problem diversity of representation in LLM generations. We present evaluation datasets and propose metrics to measure diversity in generated responses along people and culture axes. We find that LLMs understand the notion of diversity, and that they can reason and critique their own responses for that goal. This finding motivated a new prompting technique called collective-critique and self-voting (CCSV) to self-improve people diversity of LLMs by tapping into its diversity reasoning capabilities, without relying on handcrafted examples or prompt tuning. Extensive empirical experiments with both human and automated evaluations show that our proposed approach is effective at improving people and culture diversity, and outperforms all baseline methods by a large margin.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="lahoti-etal-2023-improving">
<titleInfo>
<title>Improving Diversity of Demographic Representation in Large Language Models via Collective-Critiques and Self-Voting</title>
</titleInfo>
<name type="personal">
<namePart type="given">Preethi</namePart>
<namePart type="family">Lahoti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nicholas</namePart>
<namePart type="family">Blumm</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiao</namePart>
<namePart type="family">Ma</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Raghavendra</namePart>
<namePart type="family">Kotikalapudi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sahitya</namePart>
<namePart type="family">Potluri</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Qijun</namePart>
<namePart type="family">Tan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hansa</namePart>
<namePart type="family">Srinivasan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ben</namePart>
<namePart type="family">Packer</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ahmad</namePart>
<namePart type="family">Beirami</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alex</namePart>
<namePart type="family">Beutel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jilin</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023-12</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Houda</namePart>
<namePart type="family">Bouamor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Juan</namePart>
<namePart type="family">Pino</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kalika</namePart>
<namePart type="family">Bali</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Singapore</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>A crucial challenge for generative large language models (LLMs) is diversity: when a user’s prompt is under-specified, models may follow implicit assumptions while generating a response, which may result in homogenization of the responses, as well as certain demographic groups being under-represented or even erased from the generated responses. In this paper, we formalize the problem diversity of representation in LLM generations. We present evaluation datasets and propose metrics to measure diversity in generated responses along people and culture axes. We find that LLMs understand the notion of diversity, and that they can reason and critique their own responses for that goal. This finding motivated a new prompting technique called collective-critique and self-voting (CCSV) to self-improve people diversity of LLMs by tapping into its diversity reasoning capabilities, without relying on handcrafted examples or prompt tuning. Extensive empirical experiments with both human and automated evaluations show that our proposed approach is effective at improving people and culture diversity, and outperforms all baseline methods by a large margin.</abstract>
<identifier type="citekey">lahoti-etal-2023-improving</identifier>
<identifier type="doi">10.18653/v1/2023.emnlp-main.643</identifier>
<location>
<url>https://aclanthology.org/2023.emnlp-main.643</url>
</location>
<part>
<date>2023-12</date>
<extent unit="page">
<start>10383</start>
<end>10405</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Improving Diversity of Demographic Representation in Large Language Models via Collective-Critiques and Self-Voting
%A Lahoti, Preethi
%A Blumm, Nicholas
%A Ma, Xiao
%A Kotikalapudi, Raghavendra
%A Potluri, Sahitya
%A Tan, Qijun
%A Srinivasan, Hansa
%A Packer, Ben
%A Beirami, Ahmad
%A Beutel, Alex
%A Chen, Jilin
%Y Bouamor, Houda
%Y Pino, Juan
%Y Bali, Kalika
%S Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing
%D 2023
%8 December
%I Association for Computational Linguistics
%C Singapore
%F lahoti-etal-2023-improving
%X A crucial challenge for generative large language models (LLMs) is diversity: when a user’s prompt is under-specified, models may follow implicit assumptions while generating a response, which may result in homogenization of the responses, as well as certain demographic groups being under-represented or even erased from the generated responses. In this paper, we formalize the problem diversity of representation in LLM generations. We present evaluation datasets and propose metrics to measure diversity in generated responses along people and culture axes. We find that LLMs understand the notion of diversity, and that they can reason and critique their own responses for that goal. This finding motivated a new prompting technique called collective-critique and self-voting (CCSV) to self-improve people diversity of LLMs by tapping into its diversity reasoning capabilities, without relying on handcrafted examples or prompt tuning. Extensive empirical experiments with both human and automated evaluations show that our proposed approach is effective at improving people and culture diversity, and outperforms all baseline methods by a large margin.
%R 10.18653/v1/2023.emnlp-main.643
%U https://aclanthology.org/2023.emnlp-main.643
%U https://doi.org/10.18653/v1/2023.emnlp-main.643
%P 10383-10405
Markdown (Informal)
[Improving Diversity of Demographic Representation in Large Language Models via Collective-Critiques and Self-Voting](https://aclanthology.org/2023.emnlp-main.643) (Lahoti et al., EMNLP 2023)
ACL
- Preethi Lahoti, Nicholas Blumm, Xiao Ma, Raghavendra Kotikalapudi, Sahitya Potluri, Qijun Tan, Hansa Srinivasan, Ben Packer, Ahmad Beirami, Alex Beutel, and Jilin Chen. 2023. Improving Diversity of Demographic Representation in Large Language Models via Collective-Critiques and Self-Voting. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 10383–10405, Singapore. Association for Computational Linguistics.