Awesome Auto Research Hub Papers · Datasets · Projects
← Back to datasets

AutoWorldModel-Bench

2026 Datasets
Preview for AutoWorldModel-Bench

Dataset Analysis

Introduces AutoWorldModel-Bench, a closed-loop benchmark where frontier coding agents autonomously improve world-model starters across eight game environments, showing research-style modifications dominate.

Provenance

Collected from papers.

Derived from paper: AutoWorldModel-Bench: A State-Centric Benchmark for Automated World-Model Research

Resources