Hypersim

Hypersim is a benchmark for understanding where things are in space. The official description says: "Hypersim - 3D spatial understanding benchmark using synthetic indoor scenes". Below we describe in simple words what benchmarks of this type generally examine.

Hypersim measures understanding where things are in space. In detail, it checks:

• locating and pointing to regions in images
• reasoning about positions, counts and relations between objects
• 3D-aware questions such as depth, perspective and room layout
• grounded perception tasks that must stay anchored to pixel locations

The score is accuracy: the percentage of tasks answered or solved correctly. 100% means every task was solved perfectly; higher is better.

Official methodology note: Hypersim - 3D spatial understanding benchmark using synthetic indoor scenes.

What is the Hypersim benchmark?

Hypersim is a benchmark for understanding where things are in space. The official description says: 'Hypersim - 3D spatial understanding benchmark using synthetic indoor scenes'. Below we describe in simple words what benchmarks of this type generally examine.

What does Hypersim test?

Hypersim mainly measures understanding where things are in space. Concretely, it checks: locating and pointing to regions in images; reasoning about positions, counts and relations between objects; 3D-aware questions such as depth, perspective and room layout.

How is the Hypersim score calculated?

The score is accuracy: the percentage of tasks answered or solved correctly. 100% means every task was solved perfectly; higher is better.

Where do I get Hypersim data and scores?

The official benchmark page/reference is linked on this page. Every score on Open LLM Radar is also available as open data: download the scores JSON (http://open-llm-radar-api:8000/api/v1/benchmarks/hypersim/scores) or pull it directly from the Open LLM Radar API with any HTTP client.