@inproceedings{ridoy-etal-2026-language,
title = "From Language Specifications to Executable {T}uring Machines: Evaluating {LLM}s as Computational Machine Designers",
author = "Ridoy, Shahriyar Zaman and
Hasan, S. M. Muhtasimul and
Wasi, Azmine Toushik and
Tshering, Kinga and
Tonmoy, Koushik Ahamed and
Khan, Md Mosaddek",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.2424/",
pages = "52031--52052",
ISBN = "979-8-89176-462-0",
abstract = "Recognizing a formal language does not imply the ability to construct the computation that decides it. We operationalize this distinction as an executable benchmark for language-model reasoning: given a formal-language specification, an LLM must synthesize a deterministic or nondeterministic Turing machine that decides the language. We introduce \textsc{TMEval}, a benchmark of 253 language specifications across 7 structural categories spanning context-free and non-context-free phenomena. Each language includes 600 oracle-labeled membership tests, and every output is evaluated as an executable artifact rather than a textual response. Our pipeline enforces a strict output contract, validates machine construction, executes each buildable machine, and assigns one of four outcomes: \textsc{static{\_}fail}, \textsc{build{\_}fail}, \textsc{semantic{\_}mismatch}, or \textsc{exact{\_}pass}. Across eight LLMs, $2{,}024$ model{--}language candidates, and up to $1{,}214{,}400$ behavioral evaluations, only 11.12{\%} exactly pass all tests. More than half fail before execution, over one-third implement the wrong language, and 75 languages remain unsolved by all systems. Failure depends less on the context-free/non-context-free boundary than on the operational structure required to realize the specification. \textsc{TMEval} reveals a substantial gap between formal-language recognition and machine-checked symbolic construction. \textsc{TMEval} is available at \url{https://turing-machineeval.github.io/}."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ridoy-etal-2026-language">
<titleInfo>
<title>From Language Specifications to Executable Turing Machines: Evaluating LLMs as Computational Machine Designers</title>
</titleInfo>
<name type="personal">
<namePart type="given">Shahriyar</namePart>
<namePart type="given">Zaman</namePart>
<namePart type="family">Ridoy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">S</namePart>
<namePart type="given">M</namePart>
<namePart type="given">Muhtasimul</namePart>
<namePart type="family">Hasan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Azmine</namePart>
<namePart type="given">Toushik</namePart>
<namePart type="family">Wasi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kinga</namePart>
<namePart type="family">Tshering</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Koushik</namePart>
<namePart type="given">Ahamed</namePart>
<namePart type="family">Tonmoy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Md</namePart>
<namePart type="given">Mosaddek</namePart>
<namePart type="family">Khan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Recognizing a formal language does not imply the ability to construct the computation that decides it. We operationalize this distinction as an executable benchmark for language-model reasoning: given a formal-language specification, an LLM must synthesize a deterministic or nondeterministic Turing machine that decides the language. We introduce TMEval, a benchmark of 253 language specifications across 7 structural categories spanning context-free and non-context-free phenomena. Each language includes 600 oracle-labeled membership tests, and every output is evaluated as an executable artifact rather than a textual response. Our pipeline enforces a strict output contract, validates machine construction, executes each buildable machine, and assigns one of four outcomes: static_fail, build_fail, semantic_mismatch, or exact_pass. Across eight LLMs, 2,024 model–language candidates, and up to 1,214,400 behavioral evaluations, only 11.12% exactly pass all tests. More than half fail before execution, over one-third implement the wrong language, and 75 languages remain unsolved by all systems. Failure depends less on the context-free/non-context-free boundary than on the operational structure required to realize the specification. TMEval reveals a substantial gap between formal-language recognition and machine-checked symbolic construction. TMEval is available at https://turing-machineeval.github.io/.</abstract>
<identifier type="citekey">ridoy-etal-2026-language</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.2424/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>52031</start>
<end>52052</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T From Language Specifications to Executable Turing Machines: Evaluating LLMs as Computational Machine Designers
%A Ridoy, Shahriyar Zaman
%A Hasan, S. M. Muhtasimul
%A Wasi, Azmine Toushik
%A Tshering, Kinga
%A Tonmoy, Koushik Ahamed
%A Khan, Md Mosaddek
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F ridoy-etal-2026-language
%X Recognizing a formal language does not imply the ability to construct the computation that decides it. We operationalize this distinction as an executable benchmark for language-model reasoning: given a formal-language specification, an LLM must synthesize a deterministic or nondeterministic Turing machine that decides the language. We introduce TMEval, a benchmark of 253 language specifications across 7 structural categories spanning context-free and non-context-free phenomena. Each language includes 600 oracle-labeled membership tests, and every output is evaluated as an executable artifact rather than a textual response. Our pipeline enforces a strict output contract, validates machine construction, executes each buildable machine, and assigns one of four outcomes: static_fail, build_fail, semantic_mismatch, or exact_pass. Across eight LLMs, 2,024 model–language candidates, and up to 1,214,400 behavioral evaluations, only 11.12% exactly pass all tests. More than half fail before execution, over one-third implement the wrong language, and 75 languages remain unsolved by all systems. Failure depends less on the context-free/non-context-free boundary than on the operational structure required to realize the specification. TMEval reveals a substantial gap between formal-language recognition and machine-checked symbolic construction. TMEval is available at https://turing-machineeval.github.io/.
%U https://aclanthology.org/2026.findings-emnlp.2424/
%P 52031-52052
Markdown (Informal)
[From Language Specifications to Executable Turing Machines: Evaluating LLMs as Computational Machine Designers](https://aclanthology.org/2026.findings-emnlp.2424/) (Ridoy et al., Findings 2026)
ACL