<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="del-fishel-2023-true">
    <titleInfo>
        <title>True Detective: A Deep Abductive Reasoning Benchmark Undoable for GPT-3 and Challenging for GPT-4</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Maksym</namePart>
        <namePart type="family">Del</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Mark</namePart>
        <namePart type="family">Fishel</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2023-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 12th Joint Conference on Lexical and Computational Semantics (*SEM 2023)</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Alexis</namePart>
            <namePart type="family">Palmer</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jose</namePart>
            <namePart type="family">Camacho-collados</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Toronto, Canada</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>Large language models (LLMs) have demonstrated solid zero-shot reasoning capabilities, which is reflected in their performance on the current test tasks. This calls for a more challenging benchmark requiring highly advanced reasoning ability to be solved. In this paper, we introduce such a benchmark, consisting of 191 long-form (1200 words on average) mystery narratives constructed as detective puzzles. Puzzles are sourced from the “5 Minute Mystery” platform and include a multiple-choice question for evaluation. Only 47% of humans solve a puzzle successfully on average, while the best human solvers achieve over 80% success rate. We show that GPT-3 models barely outperform random on this benchmark (with 28% accuracy) while state-of-the-art GPT-4 solves only 38% of puzzles. This indicates that there is still a significant gap in the deep reasoning abilities of LLMs and humans and highlights the need for further research in this area. Our work introduces a challenging benchmark for future studies on reasoning in language models and contributes to a better understanding of the limits of LLMs’ abilities.</abstract>
    <identifier type="citekey">del-fishel-2023-true</identifier>
    <identifier type="doi">10.18653/v1/2023.starsem-1.28</identifier>
    <location>
        <url>https://aclanthology.org/2023.starsem-1.28</url>
    </location>
    <part>
        <date>2023-07</date>
        <extent unit="page">
            <start>314</start>
            <end>322</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T True Detective: A Deep Abductive Reasoning Benchmark Undoable for GPT-3 and Challenging for GPT-4
%A Del, Maksym
%A Fishel, Mark
%Y Palmer, Alexis
%Y Camacho-collados, Jose
%S Proceedings of the 12th Joint Conference on Lexical and Computational Semantics (*SEM 2023)
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F del-fishel-2023-true
%X Large language models (LLMs) have demonstrated solid zero-shot reasoning capabilities, which is reflected in their performance on the current test tasks. This calls for a more challenging benchmark requiring highly advanced reasoning ability to be solved. In this paper, we introduce such a benchmark, consisting of 191 long-form (1200 words on average) mystery narratives constructed as detective puzzles. Puzzles are sourced from the “5 Minute Mystery” platform and include a multiple-choice question for evaluation. Only 47% of humans solve a puzzle successfully on average, while the best human solvers achieve over 80% success rate. We show that GPT-3 models barely outperform random on this benchmark (with 28% accuracy) while state-of-the-art GPT-4 solves only 38% of puzzles. This indicates that there is still a significant gap in the deep reasoning abilities of LLMs and humans and highlights the need for further research in this area. Our work introduces a challenging benchmark for future studies on reasoning in language models and contributes to a better understanding of the limits of LLMs’ abilities.
%R 10.18653/v1/2023.starsem-1.28
%U https://aclanthology.org/2023.starsem-1.28
%U https://doi.org/10.18653/v1/2023.starsem-1.28
%P 314-322

Markdown (Informal)

[True Detective: A Deep Abductive Reasoning Benchmark Undoable for GPT-3 and Challenging for GPT-4](https://aclanthology.org/2023.starsem-1.28) (Del & Fishel, *SEM 2023)

• True Detective: A Deep Abductive Reasoning Benchmark Undoable for GPT-3 and Challenging for GPT-4 (Del & Fishel, *SEM 2023)

ACL

• Maksym Del and Mark Fishel. 2023. True Detective: A Deep Abductive Reasoning Benchmark Undoable for GPT-3 and Challenging for GPT-4. In Proceedings of the 12th Joint Conference on Lexical and Computational Semantics (*SEM 2023), pages 314–322, Toronto, Canada. Association for Computational Linguistics.