Conditional Multi-Stage Failure Recovery for Embodied Agents

Youmna Farag; Svetlana Stoyanchev; Mohan Li; Simon Keizer; Rama Doddipatla

doi:10.18653/v1/2025.realm-1.15

Conditional Multi-Stage Failure Recovery for Embodied Agents

Youmna Farag, Svetlana Stoyanchev, Mohan Li, Simon Keizer, Rama Doddipatla

Correct Metadata for

Important: The Anthology treat PDFs as authoritative. Please use this form only to correct data that is out of line with the PDF. See our corrections guidelines if you need to change the PDF.

Title Adjust the title. Retain tags such as <fixed-case>.

Authors Adjust author names and order to match the PDF.

Abstract Correct abstract if needed. Retain XML formatting tags such as <tex-math>.

Verification against PDF Ensure that the new title/authors match the snapshot below. (If there is no snapshot or it is too small, consult the PDF.)

Authors concatenated from the text boxes above:

ALL author names match the snapshot above—including middle initials, hyphens, and accents.

Abstract

Embodied agents performing complex tasks are susceptible to execution failures, motivating the need for effective failure recovery mechanisms. In this work, we introduce a conditional multi-stage failure recovery framework that employs zero-shot chain prompting. The framework is structured into four error-handling stages, with three operating during task execution and one functioning as a post-execution reflection phase.Our approach utilises the reasoning capabilities of LLMs to analyse execution challenges within their environmental context and devise strategic solutions.We evaluate our method on the TfD benchmark of the TEACH dataset and achieve state-of-the-art performance, outperforming a baseline without error recovery by 11.5% and surpassing the strongest existing model by 19%.

Anthology ID:: 2025.realm-1.15
Volume:: Proceedings of the 1st Workshop for Research on Agent Language Models (REALM 2025)
Month:: July
Year:: 2025
Address:: Vienna, Austria
Editors:: Ehsan Kamalloo, Nicolas Gontier, Xing Han Lu, Nouha Dziri, Shikhar Murty, Alexandre Lacoste
Venues:: REALM | WS
SIG:
Publisher:: Association for Computational Linguistics
Note:
Pages:: 200–227
Language:
URL:: https://aclanthology.org/2025.realm-1.15/
DOI:: 10.18653/v1/2025.realm-1.15
Bibkey:
Cite (ACL):: Youmna Farag, Svetlana Stoyanchev, Mohan Li, Simon Keizer, and Rama Doddipatla. 2025. Conditional Multi-Stage Failure Recovery for Embodied Agents. In Proceedings of the 1st Workshop for Research on Agent Language Models (REALM 2025), pages 200–227, Vienna, Austria. Association for Computational Linguistics.
Cite (Informal):: Conditional Multi-Stage Failure Recovery for Embodied Agents (Farag et al., REALM 2025)
Copy Citation:
PDF:: https://aclanthology.org/2025.realm-1.15.pdf

PDF Cite Search Fix data

Export citation

BibTeX
MODS XML
Endnote
Preformatted

@inproceedings{farag-etal-2025-conditional,
    title = "Conditional Multi-Stage Failure Recovery for Embodied Agents",
    author = "Farag, Youmna  and
      Stoyanchev, Svetlana  and
      Li, Mohan  and
      Keizer, Simon  and
      Doddipatla, Rama",
    editor = "Kamalloo, Ehsan  and
      Gontier, Nicolas  and
      Lu, Xing Han  and
      Dziri, Nouha  and
      Murty, Shikhar  and
      Lacoste, Alexandre",
    booktitle = "Proceedings of the 1st Workshop for Research on Agent Language Models (REALM 2025)",
    month = jul,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2025.realm-1.15/",
    doi = "10.18653/v1/2025.realm-1.15",
    pages = "200--227",
    ISBN = "979-8-89176-264-0",
    abstract = "Embodied agents performing complex tasks are susceptible to execution failures, motivating the need for effective failure recovery mechanisms. In this work, we introduce a conditional multi-stage failure recovery framework that employs zero-shot chain prompting. The framework is structured into four error-handling stages, with three operating during task execution and one functioning as a post-execution reflection phase.Our approach utilises the reasoning capabilities of LLMs to analyse execution challenges within their environmental context and devise strategic solutions.We evaluate our method on the TfD benchmark of the TEACH dataset and achieve state-of-the-art performance, outperforming a baseline without error recovery by 11.5{\%} and surpassing the strongest existing model by 19{\%}."
}

Download as File

<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="farag-etal-2025-conditional">
    <titleInfo>
        <title>Conditional Multi-Stage Failure Recovery for Embodied Agents</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Youmna</namePart>
        <namePart type="family">Farag</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Svetlana</namePart>
        <namePart type="family">Stoyanchev</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Mohan</namePart>
        <namePart type="family">Li</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Simon</namePart>
        <namePart type="family">Keizer</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Rama</namePart>
        <namePart type="family">Doddipatla</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 1st Workshop for Research on Agent Language Models (REALM 2025)</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Ehsan</namePart>
            <namePart type="family">Kamalloo</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Nicolas</namePart>
            <namePart type="family">Gontier</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Xing</namePart>
            <namePart type="given">Han</namePart>
            <namePart type="family">Lu</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Nouha</namePart>
            <namePart type="family">Dziri</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Shikhar</namePart>
            <namePart type="family">Murty</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Alexandre</namePart>
            <namePart type="family">Lacoste</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Vienna, Austria</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-264-0</identifier>
    </relatedItem>
    <abstract>Embodied agents performing complex tasks are susceptible to execution failures, motivating the need for effective failure recovery mechanisms. In this work, we introduce a conditional multi-stage failure recovery framework that employs zero-shot chain prompting. The framework is structured into four error-handling stages, with three operating during task execution and one functioning as a post-execution reflection phase.Our approach utilises the reasoning capabilities of LLMs to analyse execution challenges within their environmental context and devise strategic solutions.We evaluate our method on the TfD benchmark of the TEACH dataset and achieve state-of-the-art performance, outperforming a baseline without error recovery by 11.5% and surpassing the strongest existing model by 19%.</abstract>
    <identifier type="citekey">farag-etal-2025-conditional</identifier>
    <identifier type="doi">10.18653/v1/2025.realm-1.15</identifier>
    <location>
        <url>https://aclanthology.org/2025.realm-1.15/</url>
    </location>
    <part>
        <date>2025-07</date>
        <extent unit="page">
            <start>200</start>
            <end>227</end>
        </extent>
    </part>
</mods>
</modsCollection>

Download as File

%0 Conference Proceedings
%T Conditional Multi-Stage Failure Recovery for Embodied Agents
%A Farag, Youmna
%A Stoyanchev, Svetlana
%A Li, Mohan
%A Keizer, Simon
%A Doddipatla, Rama
%Y Kamalloo, Ehsan
%Y Gontier, Nicolas
%Y Lu, Xing Han
%Y Dziri, Nouha
%Y Murty, Shikhar
%Y Lacoste, Alexandre
%S Proceedings of the 1st Workshop for Research on Agent Language Models (REALM 2025)
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-264-0
%F farag-etal-2025-conditional
%X Embodied agents performing complex tasks are susceptible to execution failures, motivating the need for effective failure recovery mechanisms. In this work, we introduce a conditional multi-stage failure recovery framework that employs zero-shot chain prompting. The framework is structured into four error-handling stages, with three operating during task execution and one functioning as a post-execution reflection phase.Our approach utilises the reasoning capabilities of LLMs to analyse execution challenges within their environmental context and devise strategic solutions.We evaluate our method on the TfD benchmark of the TEACH dataset and achieve state-of-the-art performance, outperforming a baseline without error recovery by 11.5% and surpassing the strongest existing model by 19%.
%R 10.18653/v1/2025.realm-1.15
%U https://aclanthology.org/2025.realm-1.15/
%U https://doi.org/10.18653/v1/2025.realm-1.15
%P 200-227

Download as File

Markdown (Informal)

[Conditional Multi-Stage Failure Recovery for Embodied Agents](https://aclanthology.org/2025.realm-1.15/) (Farag et al., REALM 2025)

Conditional Multi-Stage Failure Recovery for Embodied Agents (Farag et al., REALM 2025)

ACL

Youmna Farag, Svetlana Stoyanchev, Mohan Li, Simon Keizer, and Rama Doddipatla. 2025. Conditional Multi-Stage Failure Recovery for Embodied Agents. In Proceedings of the 1st Workshop for Research on Agent Language Models (REALM 2025), pages 200–227, Vienna, Austria. Association for Computational Linguistics.