@inproceedings{anand-shareghi-2026-cube,
title = "Cube Bench: A Controlled Diagnostic for Spatial and Sequential Reasoning in Multimodal Large Language Models",
author = "Anand, Dhruv and
Shareghi, Ehsan",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.1259/",
pages = "28678--28707",
ISBN = "979-8-89176-459-0",
abstract = "Recent multimodal large language models (MLLMs) are credited with three capabilities: they fuse vision and text, forecast the consequences of an action, and recover from mistakes through reflection. A fourth assumption underpins every agentic deployment built on these models: that strong single-step performance transfers to competent multi-step behavior. Existing evaluations cannot adjudicate these claims cleanly: static perception suites never let small errors compound, and agent suites aggregate perception, planning, and control failures into a single score. No instrument exposes where, along the perceive-forecast-act-reflect-recover loop, a multimodal model fails. We introduce Cube Bench, a controlled diagnostic built on the Rubik{'}s Cube, where transitions are deterministic and oracle distance to the goal is available at every state. Seven oracle-scored tests isolate each stage of the decision loop under matched conditions. Across open and closed-source MLLMs, current models do not reliably exhibit any of the three credited capabilities, and the fourth assumption fails most sharply: a model with frontier-adjacent single-step skill across perception, prediction, and control can sit at the floor on recovery from its own errors. This dissociation between local skill and multi-step recovery has direct implications for agentic systems that assume these capabilities compose."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="anand-shareghi-2026-cube">
<titleInfo>
<title>Cube Bench: A Controlled Diagnostic for Spatial and Sequential Reasoning in Multimodal Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Dhruv</namePart>
<namePart type="family">Anand</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ehsan</namePart>
<namePart type="family">Shareghi</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>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</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">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Recent multimodal large language models (MLLMs) are credited with three capabilities: they fuse vision and text, forecast the consequences of an action, and recover from mistakes through reflection. A fourth assumption underpins every agentic deployment built on these models: that strong single-step performance transfers to competent multi-step behavior. Existing evaluations cannot adjudicate these claims cleanly: static perception suites never let small errors compound, and agent suites aggregate perception, planning, and control failures into a single score. No instrument exposes where, along the perceive-forecast-act-reflect-recover loop, a multimodal model fails. We introduce Cube Bench, a controlled diagnostic built on the Rubik’s Cube, where transitions are deterministic and oracle distance to the goal is available at every state. Seven oracle-scored tests isolate each stage of the decision loop under matched conditions. Across open and closed-source MLLMs, current models do not reliably exhibit any of the three credited capabilities, and the fourth assumption fails most sharply: a model with frontier-adjacent single-step skill across perception, prediction, and control can sit at the floor on recovery from its own errors. This dissociation between local skill and multi-step recovery has direct implications for agentic systems that assume these capabilities compose.</abstract>
<identifier type="citekey">anand-shareghi-2026-cube</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1259/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>28678</start>
<end>28707</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Cube Bench: A Controlled Diagnostic for Spatial and Sequential Reasoning in Multimodal Large Language Models
%A Anand, Dhruv
%A Shareghi, Ehsan
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F anand-shareghi-2026-cube
%X Recent multimodal large language models (MLLMs) are credited with three capabilities: they fuse vision and text, forecast the consequences of an action, and recover from mistakes through reflection. A fourth assumption underpins every agentic deployment built on these models: that strong single-step performance transfers to competent multi-step behavior. Existing evaluations cannot adjudicate these claims cleanly: static perception suites never let small errors compound, and agent suites aggregate perception, planning, and control failures into a single score. No instrument exposes where, along the perceive-forecast-act-reflect-recover loop, a multimodal model fails. We introduce Cube Bench, a controlled diagnostic built on the Rubik’s Cube, where transitions are deterministic and oracle distance to the goal is available at every state. Seven oracle-scored tests isolate each stage of the decision loop under matched conditions. Across open and closed-source MLLMs, current models do not reliably exhibit any of the three credited capabilities, and the fourth assumption fails most sharply: a model with frontier-adjacent single-step skill across perception, prediction, and control can sit at the floor on recovery from its own errors. This dissociation between local skill and multi-step recovery has direct implications for agentic systems that assume these capabilities compose.
%U https://aclanthology.org/2026.emnlp-main.1259/
%P 28678-28707
Markdown (Informal)
[Cube Bench: A Controlled Diagnostic for Spatial and Sequential Reasoning in Multimodal Large Language Models](https://aclanthology.org/2026.emnlp-main.1259/) (Anand & Shareghi, EMNLP 2026)
ACL