Benchmark RadarAI BENCHMARK PROFILE
SIMPLE
SIMPLE is a simulation testbed for humanoid loco-manipulation with 60 tasks, 50 scenes, and over 1,000 objects, integrating data generation pipelines and benchmarking policies.
- Released
- 2026-06-06
- Readiness
- Paper only
- Primary field
- Robotics & Autonomous Systems
Why it matters
Real-world humanoid evaluation is expensive; SIMPLE aims to provide a reproducible simulation benchmark, but public access to environments and data is not yet confirmed.
Motivation
Humanoid foundation models are advancing faster than we can evaluate them.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.