Benchmark RadarAI BENCHMARK PROFILE
The Imitator Game
Evaluates robot imitation across four levels of increasing divergence between demonstration and robot scene, plus an open platform for blind A/B human evaluation.
- Released
- 2026-08-23
- Readiness
- Paper only
- Primary field
- Robotics & Autonomous Systems
Why it matters
Identifies functional substitution as the barrier to intent-level imitation, moving evaluation beyond trajectory replay toward task understanding.
Motivation
Humans imitate at the level of intent: given a demonstration, we infer its goal and carry it out with whatever tools, objects, and layouts are at hand.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.