@inproceedings{wang-etal-2026-architect,
title = "{ARCHITECT}: Uncertainty-Aware Dynamic Tool Learning via Causal Intervention for Open-World Agents",
author = "Wang, Zhangyi and
Xu, Jiexiang and
Yu, Bingnan and
Li, Zongze",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.acl-long.1707/",
doi = "10.18653/v1/2026.acl-long.1707",
pages = "36819--36829",
ISBN = "979-8-89176-390-6",
abstract = "Dynamic tool generation empowers Large Language Model (LLM) agents to synthesize tools on demand, yet a critical challenge remains: 32.4{\%} of generated tools fail on first invocation. We present Causal Tool Diagnosis (CTD), a principled framework that moves beyond black-box reliability prediction to interpretable failure attribution. CTD constructs a Structural Causal Model (SCM) capturing how specification quality, code characteristics, and execution environment jointly determine tool outcomes. Uniquely leveraging code{'}s intervenability, we conduct controlled sandbox experiments to estimate causal effects{---}an advantage unavailable in pure text generation. CTD jointly predicts confidence (Spearman rank correlation coefficient $\rho$=0.90) and root cause attribution (78{\%} accuracy), with attributions directly guiding targeted repairs (+9.6{\%} success rate over error-type classification). Our ARCHITECT framework, integrating CTD throughout the tool lifecycle, achieves state-of-the-art on four benchmarks including StableToolBench (+3.8{\%}), MINT (+4.6{\%}), T-Eval (+3.7{\%}), and SWE-bench Lite (+2.4{\%}), with consistent improvements across all settings."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2026-architect">
<titleInfo>
<title>ARCHITECT: Uncertainty-Aware Dynamic Tool Learning via Causal Intervention for Open-World Agents</title>
</titleInfo>
<name type="personal">
<namePart type="given">Zhangyi</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiexiang</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bingnan</namePart>
<namePart type="family">Yu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zongze</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-390-6</identifier>
</relatedItem>
<abstract>Dynamic tool generation empowers Large Language Model (LLM) agents to synthesize tools on demand, yet a critical challenge remains: 32.4% of generated tools fail on first invocation. We present Causal Tool Diagnosis (CTD), a principled framework that moves beyond black-box reliability prediction to interpretable failure attribution. CTD constructs a Structural Causal Model (SCM) capturing how specification quality, code characteristics, and execution environment jointly determine tool outcomes. Uniquely leveraging code’s intervenability, we conduct controlled sandbox experiments to estimate causal effects—an advantage unavailable in pure text generation. CTD jointly predicts confidence (Spearman rank correlation coefficient ρ=0.90) and root cause attribution (78% accuracy), with attributions directly guiding targeted repairs (+9.6% success rate over error-type classification). Our ARCHITECT framework, integrating CTD throughout the tool lifecycle, achieves state-of-the-art on four benchmarks including StableToolBench (+3.8%), MINT (+4.6%), T-Eval (+3.7%), and SWE-bench Lite (+2.4%), with consistent improvements across all settings.</abstract>
<identifier type="citekey">wang-etal-2026-architect</identifier>
<identifier type="doi">10.18653/v1/2026.acl-long.1707</identifier>
<location>
<url>https://aclanthology.org/2026.acl-long.1707/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>36819</start>
<end>36829</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T ARCHITECT: Uncertainty-Aware Dynamic Tool Learning via Causal Intervention for Open-World Agents
%A Wang, Zhangyi
%A Xu, Jiexiang
%A Yu, Bingnan
%A Li, Zongze
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-390-6
%F wang-etal-2026-architect
%X Dynamic tool generation empowers Large Language Model (LLM) agents to synthesize tools on demand, yet a critical challenge remains: 32.4% of generated tools fail on first invocation. We present Causal Tool Diagnosis (CTD), a principled framework that moves beyond black-box reliability prediction to interpretable failure attribution. CTD constructs a Structural Causal Model (SCM) capturing how specification quality, code characteristics, and execution environment jointly determine tool outcomes. Uniquely leveraging code’s intervenability, we conduct controlled sandbox experiments to estimate causal effects—an advantage unavailable in pure text generation. CTD jointly predicts confidence (Spearman rank correlation coefficient ρ=0.90) and root cause attribution (78% accuracy), with attributions directly guiding targeted repairs (+9.6% success rate over error-type classification). Our ARCHITECT framework, integrating CTD throughout the tool lifecycle, achieves state-of-the-art on four benchmarks including StableToolBench (+3.8%), MINT (+4.6%), T-Eval (+3.7%), and SWE-bench Lite (+2.4%), with consistent improvements across all settings.
%R 10.18653/v1/2026.acl-long.1707
%U https://aclanthology.org/2026.acl-long.1707/
%U https://doi.org/10.18653/v1/2026.acl-long.1707
%P 36819-36829
Markdown (Informal)
[ARCHITECT: Uncertainty-Aware Dynamic Tool Learning via Causal Intervention for Open-World Agents](https://aclanthology.org/2026.acl-long.1707/) (Wang et al., ACL 2026)
ACL