List your company

ResearchFormerly WizWand

Autonomous Driving Evaluation

Benchmarks

DatasetSOTA methodMetricTrendResultsLast updated

DriveLM VRU-Accident benchmark

64.13GAR

10

May 12, 2026

Dolphins VRU-Accident benchmark

89.33GAR

5

May 12, 2026

nuPlan Generated safety-critical scenarios

50Collision Rate (%)

4

Feb 26, 2026

nuPlan Original scenarios

8.3CR

4

Feb 26, 2026

Composite held-out set (Switzerland)

w/o fine-tuning

95.4CR

3

Jun 30, 2026

Composite Small Towns (held-out set)

w/o fine-tuning

95.6CR

3

Jun 30, 2026

Composite Yosemite (held-out set)

w/o fine-tuning

96.5CR

3

Jun 30, 2026

Composite held-out set (Yellowstone)

w/o fine-tuning

95.3CR

3

Jun 30, 2026

Composite (New York City)

w/o fine-tuning

97.3Completion Rate (CR)

3

Jun 30, 2026

Composite Los Angeles (held-out set)

w/o fine-tuning

96.8CR

3

Jun 30, 2026