@inproceedings{vu-etal-2024-foundational,
title = "Foundational Autoraters: Taming Large Language Models for Better Automatic Evaluation",
author = "Vu, Tu and
Krishna, Kalpesh and
Alzubi, Salaheddin and
Tar, Chris and
Faruqui, Manaal and
Sung, Yun-Hsuan",
editor = "Al-Onaizan, Yaser and
Bansal, Mohit and
Chen, Yun-Nung",
booktitle = "Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing",
month = nov,
year = "2024",
address = "Miami, Florida, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.emnlp-main.949",
pages = "17086--17105",
abstract = "As large language models (LLMs) evolve, evaluating their output reliably becomes increasingly difficult due to the high cost of human evaluation. To address this, we introduce FLAMe, a family of Foundational Large Autorater Models. FLAMe is trained on a diverse set of over 100 quality assessment tasks, incorporating 5M+ human judgments curated from publicly released human evaluations. FLAMe outperforms models like GPT-4 and Claude-3 on various held-out tasks, and serves as a powerful starting point for fine-tuning, as shown in our reward model evaluation case study (FLAMe-RM). On Reward-Bench, FLAMe-RM-24B achieves 87.8{\%} accuracy, surpassing GPT-4-0125 (85.9{\%}) and GPT-4o (84.7{\%}). Additionally, we introduce FLAMe-Opt-RM, an efficient tail-patch fine-tuning approach that offers competitive RewardBench performance using 25{\mbox{$\times$}}fewer training datapoints. Our FLAMe variants outperform popular proprietary LLM-as-a-Judge models on 8 of 12 autorater benchmarks, covering 53 quality assessment tasks, including RewardBench and LLM-AggreFact. Finally, our analysis shows that FLAMe is significantly less biased than other LLM-as-a-Judge models on the CoBBLEr autorater bias benchmark.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="vu-etal-2024-foundational">
<titleInfo>
<title>Foundational Autoraters: Taming Large Language Models for Better Automatic Evaluation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Tu</namePart>
<namePart type="family">Vu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kalpesh</namePart>
<namePart type="family">Krishna</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Salaheddin</namePart>
<namePart type="family">Alzubi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chris</namePart>
<namePart type="family">Tar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Manaal</namePart>
<namePart type="family">Faruqui</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yun-Hsuan</namePart>
<namePart type="family">Sung</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>Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing</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>As large language models (LLMs) evolve, evaluating their output reliably becomes increasingly difficult due to the high cost of human evaluation. To address this, we introduce FLAMe, a family of Foundational Large Autorater Models. FLAMe is trained on a diverse set of over 100 quality assessment tasks, incorporating 5M+ human judgments curated from publicly released human evaluations. FLAMe outperforms models like GPT-4 and Claude-3 on various held-out tasks, and serves as a powerful starting point for fine-tuning, as shown in our reward model evaluation case study (FLAMe-RM). On Reward-Bench, FLAMe-RM-24B achieves 87.8% accuracy, surpassing GPT-4-0125 (85.9%) and GPT-4o (84.7%). Additionally, we introduce FLAMe-Opt-RM, an efficient tail-patch fine-tuning approach that offers competitive RewardBench performance using 25\timesfewer training datapoints. Our FLAMe variants outperform popular proprietary LLM-as-a-Judge models on 8 of 12 autorater benchmarks, covering 53 quality assessment tasks, including RewardBench and LLM-AggreFact. Finally, our analysis shows that FLAMe is significantly less biased than other LLM-as-a-Judge models on the CoBBLEr autorater bias benchmark.</abstract>
<identifier type="citekey">vu-etal-2024-foundational</identifier>
<location>
<url>https://aclanthology.org/2024.emnlp-main.949</url>
</location>
<part>
<date>2024-11</date>
<extent unit="page">
<start>17086</start>
<end>17105</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Foundational Autoraters: Taming Large Language Models for Better Automatic Evaluation
%A Vu, Tu
%A Krishna, Kalpesh
%A Alzubi, Salaheddin
%A Tar, Chris
%A Faruqui, Manaal
%A Sung, Yun-Hsuan
%Y Al-Onaizan, Yaser
%Y Bansal, Mohit
%Y Chen, Yun-Nung
%S Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing
%D 2024
%8 November
%I Association for Computational Linguistics
%C Miami, Florida, USA
%F vu-etal-2024-foundational
%X As large language models (LLMs) evolve, evaluating their output reliably becomes increasingly difficult due to the high cost of human evaluation. To address this, we introduce FLAMe, a family of Foundational Large Autorater Models. FLAMe is trained on a diverse set of over 100 quality assessment tasks, incorporating 5M+ human judgments curated from publicly released human evaluations. FLAMe outperforms models like GPT-4 and Claude-3 on various held-out tasks, and serves as a powerful starting point for fine-tuning, as shown in our reward model evaluation case study (FLAMe-RM). On Reward-Bench, FLAMe-RM-24B achieves 87.8% accuracy, surpassing GPT-4-0125 (85.9%) and GPT-4o (84.7%). Additionally, we introduce FLAMe-Opt-RM, an efficient tail-patch fine-tuning approach that offers competitive RewardBench performance using 25\timesfewer training datapoints. Our FLAMe variants outperform popular proprietary LLM-as-a-Judge models on 8 of 12 autorater benchmarks, covering 53 quality assessment tasks, including RewardBench and LLM-AggreFact. Finally, our analysis shows that FLAMe is significantly less biased than other LLM-as-a-Judge models on the CoBBLEr autorater bias benchmark.
%U https://aclanthology.org/2024.emnlp-main.949
%P 17086-17105
Markdown (Informal)
[Foundational Autoraters: Taming Large Language Models for Better Automatic Evaluation](https://aclanthology.org/2024.emnlp-main.949) (Vu et al., EMNLP 2024)
ACL