@inproceedings{efrat-etal-2021-cryptonite,
title = "Cryptonite: A Cryptic Crossword Benchmark for Extreme Ambiguity in Language",
author = "Efrat, Avia and
Shaham, Uri and
Kilman, Dan and
Levy, Omer",
booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing",
month = nov,
year = "2021",
address = "Online and Punta Cana, Dominican Republic",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.emnlp-main.344",
doi = "10.18653/v1/2021.emnlp-main.344",
pages = "4186--4192",
abstract = "Current NLP datasets targeting ambiguity can be solved by a native speaker with relative ease. We present Cryptonite, a large-scale dataset based on cryptic crosswords, which is both linguistically complex and naturally sourced. Each example in Cryptonite is a cryptic clue, a short phrase or sentence with a misleading surface reading, whose solving requires disambiguating semantic, syntactic, and phonetic wordplays, as well as world knowledge. Cryptic clues pose a challenge even for experienced solvers, though top-tier experts can solve them with almost 100{\%} accuracy. Cryptonite is a challenging task for current models; fine-tuning T5-Large on 470k cryptic clues achieves only 7.6{\%} accuracy, on par with the accuracy of a rule-based clue solver (8.6{\%}).",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="efrat-etal-2021-cryptonite">
<titleInfo>
<title>Cryptonite: A Cryptic Crossword Benchmark for Extreme Ambiguity in Language</title>
</titleInfo>
<name type="personal">
<namePart type="given">Avia</namePart>
<namePart type="family">Efrat</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Uri</namePart>
<namePart type="family">Shaham</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dan</namePart>
<namePart type="family">Kilman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Omer</namePart>
<namePart type="family">Levy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2021-nov</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Online and Punta Cana, Dominican Republic</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Current NLP datasets targeting ambiguity can be solved by a native speaker with relative ease. We present Cryptonite, a large-scale dataset based on cryptic crosswords, which is both linguistically complex and naturally sourced. Each example in Cryptonite is a cryptic clue, a short phrase or sentence with a misleading surface reading, whose solving requires disambiguating semantic, syntactic, and phonetic wordplays, as well as world knowledge. Cryptic clues pose a challenge even for experienced solvers, though top-tier experts can solve them with almost 100% accuracy. Cryptonite is a challenging task for current models; fine-tuning T5-Large on 470k cryptic clues achieves only 7.6% accuracy, on par with the accuracy of a rule-based clue solver (8.6%).</abstract>
<identifier type="citekey">efrat-etal-2021-cryptonite</identifier>
<identifier type="doi">10.18653/v1/2021.emnlp-main.344</identifier>
<location>
<url>https://aclanthology.org/2021.emnlp-main.344</url>
</location>
<part>
<date>2021-nov</date>
<extent unit="page">
<start>4186</start>
<end>4192</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Cryptonite: A Cryptic Crossword Benchmark for Extreme Ambiguity in Language
%A Efrat, Avia
%A Shaham, Uri
%A Kilman, Dan
%A Levy, Omer
%S Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing
%D 2021
%8 nov
%I Association for Computational Linguistics
%C Online and Punta Cana, Dominican Republic
%F efrat-etal-2021-cryptonite
%X Current NLP datasets targeting ambiguity can be solved by a native speaker with relative ease. We present Cryptonite, a large-scale dataset based on cryptic crosswords, which is both linguistically complex and naturally sourced. Each example in Cryptonite is a cryptic clue, a short phrase or sentence with a misleading surface reading, whose solving requires disambiguating semantic, syntactic, and phonetic wordplays, as well as world knowledge. Cryptic clues pose a challenge even for experienced solvers, though top-tier experts can solve them with almost 100% accuracy. Cryptonite is a challenging task for current models; fine-tuning T5-Large on 470k cryptic clues achieves only 7.6% accuracy, on par with the accuracy of a rule-based clue solver (8.6%).
%R 10.18653/v1/2021.emnlp-main.344
%U https://aclanthology.org/2021.emnlp-main.344
%U https://doi.org/10.18653/v1/2021.emnlp-main.344
%P 4186-4192
Markdown (Informal)
[Cryptonite: A Cryptic Crossword Benchmark for Extreme Ambiguity in Language](https://aclanthology.org/2021.emnlp-main.344) (Efrat et al., EMNLP 2021)
ACL