Embodied Playground: A Long-Horizon Agentic Robotics Benchmark for Physical, Strategic, and Spatial Reasoning
- Published
- Source
- arXiv
- Paper number
- 1371
- Field
- Robotics
- arXiv ID
- 2610.27039
Key points
- The playground comprises 410 task environments spanning basic manipulation, sequential puzzles, and multi-stage household activities.
- The authors curated EP-100, a representative selection of 100 tasks across five suites and 16 task types, together with the EP-Easy and EP-Hard difficulty splits and the compact EP-20 selection.
- GPT-6-Astra achieves 13.80% Pass@1 and 19.94% Pass@2 on EP-100.
- Opus-5.5-extra achieves 29.22% Pass@1 and 35.85% Pass@2 on the same evaluation.
- For GPT-6-Astra, EP-Easy reaches 18.42% Pass@1 and 27.41% Pass@2, compared with 8.81% and 12.01% on EP-Hard.
- The audit identifies incomplete rule or threshold disclosure affecting 574 of the 645 scored failures (88.99%).
Paper links
External research summaries. These are not HDATF publications or measured product results.