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

Transformers and Slot Encoding for Sample Efficient Physical World Modelling

arXiv 24.5 2024 66.5 method

TLDR

Combines Transformers with slot-attention for sample-efficient world modelling from video, improving over image-level approaches.

Reasoning

The paper proposes a novel architecture integrating slot-attention to model object-level representations, showing improved sample efficiency and reduced performance variance. Strengths include clear methodology and available code; weaknesses include lack of real-world validation and limited scope to video input.

Read-first score

Read-first score 66.5, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 42.

Reproducibility 25%
81

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

Recency 8%
75.1

Uses a gentle age decay so recent papers surface without erasing older foundations. 2024

Topical relevance 42%
60

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 25%
60

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

Field roles

Reproducibility anchor

Rank sensitivity

Stability: volatile; rank range: 198.

Keyword Scores

world model
10
video world model
9
world dynamics prediction
8
generative world model
6
model-based reinforcement learning world model
4
world simulator
3
interactive world model
2

Deep Analysis

Innovations

  • Combining Transformers with slot-attention for world modelling from video input
  • Learning object-level representations for world modelling instead of image-level
  • Improved sample efficiency and reduced performance variation compared to existing solutions

Methodology

The proposed architecture integrates Transformers for world modelling with the slot-attention paradigm to learn object representations from video input. The model is trained and evaluated on video-based world modelling tasks, comparing sample efficiency and performance variation against existing image-level Transformer approaches.

Key Results

The architecture demonstrates improved sample efficiency over existing solutions and a reduction in the variation of performance across training examples.

Resources

PaperCode

Tags