Curiosity-Driven Exploration by Self-Supervised Prediction

arXiv 2017 20.2 method

TLDR

Curiosity as prediction error in a self-supervised learned feature space enables exploration in sparse reward environments.

Reasoning

The paper introduces a novel curiosity formulation that scales to high-dimensional states and ignores irrelevant features, but its evaluation is limited to two game environments and lacks comparisons to other exploration methods.

Read-first score

Read-first score 20.2, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 9.

Reproducibility 25%

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

Recency 8%

27.6

Uses a gentle age decay so recent papers surface without erasing older foundations. 2017

Methodology quality 25%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=none

Topical relevance 42%

12.9

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

Field roles

Candidate

Rank sensitivity

Stability: volatile; rank range: 17.

Keyword Scores

world dynamics prediction

world model

model-based reinforcement learning world model

world simulator

generative world model

interactive world model

video world model

Resources

Paper