GUI-CC: Benchmarking Contextual Consistency of GUI World Models as Agent Environments

arXiv 2026 40.3 benchmark

TLDR

Introduces GUI-CC, a benchmark evaluating whether GUI world models maintain contextual consistency in multi-step agent environments, showing single-step accuracy does not guarantee reliable rollouts.

Reasoning

Strengths: addresses a key gap between one-step evaluation and multi-step use, with offline and online tracks using real mobile trajectories and emulator-verified tasks. Weaknesses: visible abstract lacks details on baselines and metric formulations, and the benchmark is limited to mobile GUI environments.

Read-first score

Read-first score 40.3, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 48.

Recency 5%

100

Uses a gentle age decay so recent papers surface without erasing older foundations. 2026

Methodology quality 15%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=benchmark,evaluation,experiment,metric

Topical relevance 25%

68.6

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Reproducibility 15%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=none

Graph prestige 15%

10.9

PageRank-style prestige over the local citation graph built from available OpenAlex references. local_pagerank=0.0031

Citation impact 15%

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. cited_by_count=0

Citation velocity 10%

Citation velocity estimates citations per publication-year to reduce old-paper bias. velocity=0.00

Field roles

FrontierMethodology anchor

Rank sensitivity

Stability: volatile; rank range: 410.

Keyword Scores

world model

world simulator

generative world model

interactive world model

world dynamics prediction

model-based reinforcement learning world model

video world model

Deep Analysis

Innovations

Introduces GUI-CC, a benchmark for evaluating contextual consistency of GUI world models as multi-step agent environments rather than isolated one-step next-screen predictors.
Provides two complementary evaluation tracks: an offline reference-action track that rolls models along real mobile GUI trajectories, and an online agent-loop track where fixed probing agents interact with model-generated UIs.
Constructs 500 offline trajectory tasks from GUIOdyssey and 200 emulator-verified online tasks across 30 mobile apps, with metrics for transition fidelity, transition plausibility, contextual consistency, and task progress.

Methodology

GUI-CC evaluates GUI world models through offline rollouts along reference mobile GUI trajectories and online agent-loop interactions with model-generated UIs. It includes 500 offline tasks derived from GUIOdyssey and 200 emulator-verified online tasks across 30 mobile apps, measuring transition fidelity, transition plausibility, contextual consistency, and task progress.

Key Results

Plausible single-step generation does not guarantee reliable environment simulation: current models often produce usable-looking screens while failing to preserve task-relevant context or support executable multi-step rollouts.

Resources

Paper