CURIO: Curiosity-Driven Test-Time Learning for Open-Ended Discovery

arXiv 2026 22.1 method

TLDR

CURIO augments test-time RL with an intrinsic curiosity world model, improving open-ended discovery on math and single-cell denoising tasks.

Reasoning

The paper presents a clear framework and strong empirical results across tasks and model scales, including diversity metrics. However, its world model is limited to hidden-state transition prediction for curiosity bonuses, not a full generative or interactive world simulator, and the abstract omits explicit limitations.

Read-first score

Read-first score 22.1, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 21.

Recency 5%

100

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

Reproducibility 15%

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

Topical relevance 25%

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 15%

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

Graph prestige 15%

5.7

PageRank-style prestige over the local citation graph built from available OpenAlex references. local_pagerank=0.0016

Citation impact 15%

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

Citation velocity 10%

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

Field roles

Frontier

Rank sensitivity

Stability: volatile; rank range: 158.

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