EA-WM: Event-Aware Generative World Model with Structured Kinematic-to-Visual Action Fields

arXiv 2026 57.3 method

TLDR

EA-WM uses structured kinematic-to-visual action fields and event-aware fusion to improve video generation for robotic world models, achieving SOTA on WorldArena.

Reasoning

The paper introduces a novel method to bridge kinematic control and visual perception for generative world models, with strong empirical results on a benchmark. However, it lacks explicit real-world robot validation and does not address model-based RL directly.

Read-first score

Read-first score 57.3, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 50.

Recency 6%

100

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

Citation impact 18%

72.6

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. citation_normalized_percentile=0.72639911

Topical relevance 29%

71.4

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

Methodology quality 18%

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

Reproducibility 18%

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

Citation velocity 12%

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

Field roles

FrontierBridgeMethodology anchor

Rank sensitivity

Stability: volatile; rank range: 385.

Keyword Scores

world model

generative world model

video world model

world dynamics prediction

world simulator

interactive world model

model-based reinforcement learning world model

Deep Analysis

Innovations

Structured Kinematic-to-Visual Action Fields that project actions and kinematic states directly into the target camera view as a geometrically grounded representation
Event-aware bidirectional fusion blocks that modulate cross-branch attention to capture object state changes and interaction dynamics

Methodology

EA-WM leverages pretrained video diffusion models as a foundation. It projects actions and kinematic states into the target camera view as Structured Kinematic-to-Visual Action Fields, then uses event-aware bidirectional fusion blocks to modulate cross-branch attention, capturing object state changes and interaction dynamics. The model is evaluated on the WorldArena benchmark.

Key Results

EA-WM achieves state-of-the-art performance on the WorldArena benchmark, outperforming existing baselines by a significant margin.

Resources

Paper