List your company

ResearchFormerly WizWand

Human Evaluation of Explanations

Benchmarks

DatasetSOTA methodMetricTrendResultsLast updated

2.15Helpfulness (MAR)

5

Apr 28, 2026

1.917Helpfulness (MAR)

5

Apr 28, 2026

DRAMA-Reasoning (val)

74.7Usefulness (UniDrive Win %)

1

Jun 24, 2026