Rushes Benchmark

What is Rushes?

A dataset and diagnostic benchmark of 44,226 user decisions in AI-generated branching narratives. Models predict each user's choice at the event level.

Released

2026-07-22

Evaluates

General AI, Language & Knowledge, Safety

Openness

unknown

Importer

Claire Radar

Review status

ai-reviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2607.20767

Rushes paper and code

• Paper https://arxiv.org/abs/2607.20767
• Code https://github.com/microsoft/rushes

Which sources cite Rushes?

• Rushes

Related General AI, Language & Knowledge, Safety benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 3DCodeBench
• 5-Dialects-BN
• A Large-Scale Dataset and Benchmark
• A3S-Bench
• AARRI-Bench
• ABC-Bench
• ActBench