World-Ego Modeling for Long-Horizon Evolution in Hybrid Embodied Tasks

arXiv 2026 61.6 method

TLDR

Introduces World-Ego Modeling to decompose future evolution into world and ego components, with a new benchmark and model achieving SOTA.

Reasoning

Strengths include a novel conceptual paradigm, a new benchmark (HTEWorld) for hybrid tasks, and strong empirical results. Weaknesses are the lack of real-world validation and potential overfitting to the specific benchmark.

Read-first score

Read-first score 61.6, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 52.

Recency 6%

100

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

Citation impact 18%

82.1

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. citation_normalized_percentile=0.82096218

Methodology quality 18%

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

Topical relevance 29%

74.3

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 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

FoundationFrontierBridgeMethodology anchor

Rank sensitivity

Stability: volatile; rank range: 421.

Keyword Scores

world model

world dynamics prediction

generative world model

world simulator

video world model

interactive world model

model-based reinforcement learning world model

Deep Analysis

Innovations

World-Ego Modeling paradigm that decomposes future evolution into world and ego components
Definition of world-ego boundary from motion-, semantic-, and intention-based perspectives
Analysis of three disentanglement strategies: post-, pre-, and full disentanglement
World-Ego Model (WEM) with an implicit separate world-ego planner and cascade-parallel mixture-of-experts (CP-MoE) diffusion generator
HTEWorld benchmark: first long-horizon world modeling benchmark for hybrid navigation-manipulation tasks with 125K video clips and 300 multi-turn evaluation trajectories

Methodology

The paper introduces World-Ego Modeling, a paradigm that separates future evolution into world (instruction-agnostic scene regularities) and ego (robot-centric instruction-conditioned dynamics) components. It instantiates this as the World-Ego Model (WEM), which couples an implicit separate world-ego planner with a cascade-parallel mixture-of-experts (CP-MoE) diffusion generator. To evaluate, the authors construct HTEWorld, a benchmark containing 125K video clips (over 4.5M frames) with fine-grained action annotations and 300 multi-turn evaluation trajectories (over 2K instructions) for hybrid navigation-manipulation tasks.

Key Results

WEM achieves state-of-the-art performance on the HTEWorld benchmark while remaining competitive on existing manipulation-only benchmarks.

Resources

Paper