Iconicity in large language models
Identifikátory výsledku
Kód výsledku v IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11210%2F25%3A10507824" target="_blank" >RIV/00216208:11210/25:10507824 - isvavai.cz</a>
Nalezeny alternativní kódy
RIV/00216208:11320/26:UUGWL4F6 RIV/61989592:15210/25:73635953
Výsledek na webu
<a href="https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=qqTCHvc.qc" target="_blank" >https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=qqTCHvc.qc</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1093/llc/fqaf095" target="_blank" >10.1093/llc/fqaf095</a>
Alternativní jazyky
Jazyk výsledku
angličtina
Název v původním jazyce
Iconicity in large language models
Popis výsledku v původním jazyce
Lexical iconicity, a direct relation between a word's meaning and its form, is an important aspect of every natural language, most commonly manifesting through sound-meaning associations. Since Large language models' (LLMs') access to both meaning and sound of text is only mediated (meaning through textual context, sound through written representation, further complicated by tokenization), we might expect that the encoding of iconicity in LLMs would be either insufficient or significantly different from human processing. This study addresses this hypothesis by having GPT-4 generate highly iconic pseudowords in artificial languages. To verify that these words actually carry iconicity, we had their meanings guessed by Czech and German participants (n = 672) and subsequently by LLM-based participants (generated by GPT-4 and Claude 3.5 Sonnet). The results revealed that humans can guess the meanings of pseudowords in the generated iconic language more accurately than words in distant natural languages and that LLM-based participants are even more successful than humans in this task. This core finding is accompanied by several additional analyses concerning the universality of the generated language and the cues that both human and LLM-based participants utilize.
Název v anglickém jazyce
Iconicity in large language models
Popis výsledku anglicky
Lexical iconicity, a direct relation between a word's meaning and its form, is an important aspect of every natural language, most commonly manifesting through sound-meaning associations. Since Large language models' (LLMs') access to both meaning and sound of text is only mediated (meaning through textual context, sound through written representation, further complicated by tokenization), we might expect that the encoding of iconicity in LLMs would be either insufficient or significantly different from human processing. This study addresses this hypothesis by having GPT-4 generate highly iconic pseudowords in artificial languages. To verify that these words actually carry iconicity, we had their meanings guessed by Czech and German participants (n = 672) and subsequently by LLM-based participants (generated by GPT-4 and Claude 3.5 Sonnet). The results revealed that humans can guess the meanings of pseudowords in the generated iconic language more accurately than words in distant natural languages and that LLM-based participants are even more successful than humans in this task. This core finding is accompanied by several additional analyses concerning the universality of the generated language and the cues that both human and LLM-based participants utilize.
Klasifikace
Druh
J<sub>imp</sub> - Článek v periodiku v databázi Web of Science
CEP obor
—
OECD FORD obor
60203 - Linguistics
Návaznosti výsledku
Projekt
<a href="/cs/project/GA24-11725S" target="_blank" >GA24-11725S: Velké jazykové modely prizmatem korpusové lingvistiky</a><br>
Návaznosti
P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)
Ostatní
Rok uplatnění
2025
Kód důvěrnosti údajů
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Údaje specifické pro druh výsledku
Název periodika
Digital Scholarship in the Humanities
ISSN
2055-7671
e-ISSN
2055-768X
Svazek periodika
40
Číslo periodika v rámci svazku
4
Stát vydavatele periodika
US - Spojené státy americké
Počet stran výsledku
22
Strana od-do
1203-1224
Kód UT WoS článku
001574575700001
EID výsledku v databázi Scopus
2-s2.0-105020300683