⌁ Awesome World Model Hub Papers · Datasets · Projects
中文 GitHub
← Back to papers

Large Emotional World Model

arXiv 25.12 2025 38.6 method, benchmark

TLDR

Proposes Large Emotional World Model integrating emotion into world models, improving prediction of emotion-driven social behaviors.

Reasoning

Strengths include novel integration of emotion into world models and empirical validation with a constructed dataset. Weaknesses are limited scope to social behaviors and lack of comparison to other world model types beyond basic tasks.

Read-first score

Read-first score 38.6, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 26.

Methodology quality 18%
90

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

Recency 6%
86.7

Uses a gentle age decay so recent papers surface without erasing older foundations. 2025

Reproducibility 18%
38

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=dataset

Topical relevance 29%
37.1

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

Citation impact 18%
0

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

Citation velocity 12%
0

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: 398.

Keyword Scores

world model
9
world dynamics prediction
8
world simulator
3
generative world model
2
model-based reinforcement learning world model
2
interactive world model
1
video world model
1

Deep Analysis

Innovations

  • Demonstrating that removing emotionally relevant information degrades reasoning performance, highlighting the importance of emotion in world understanding.
  • Proposing the Large Emotional World Model (LEWM) inspired by theory of mind, which explicitly models emotional states alongside visual observations and actions.
  • Constructing the Emotion-Why-How (EWH) dataset that integrates emotion into causal relationships, enabling reasoning about why actions occur and how emotions drive future world states.
  • Enabling the world model to predict both future states and emotional transitions.

Methodology

The authors first construct the Emotion-Why-How (EWH) dataset, which integrates emotion into causal relationships. Then they propose LEWM, which explicitly models emotional states alongside visual observations and actions, allowing the world model to predict both future states and emotional transitions. The model is evaluated on emotion-driven social behaviors and compared to general world models on basic tasks.

Key Results

LEWM more accurately predicts emotion-driven social behaviors while maintaining comparable performance to general world models on basic tasks.

Limitations

  • The EWH dataset may not capture the full complexity and diversity of real-world emotional contexts.
  • The model's performance on basic tasks is only comparable, not superior, to general world models, suggesting limited improvement in non-emotional domains.
  • The evaluation is focused on emotion-driven social behaviors, leaving broader world modeling capabilities unverified.
  • The abstract does not discuss generalization to other modalities or real-world deployment challenges.

Resources

Paper

Tags