Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Human Evaluation of Explanations on DRAMA-Reasoning (val)
Loading...
74.7
Usefulness (UniDrive Win %)
UniDrive
70.965
72.8325
74.7
76.5675
Jun 23, 2026
Usefulness (UniDrive Win %)
Usefulness (Baseline Win %)
Usefulness (Tie / No Pref. %)
Accuracy (UniDrive Win %)
Accuracy (Baseline Win %)
Accuracy (Tie / No Pref. %)
Trustworthiness (UniDrive Win %)
Trustworthiness (Baseline Win %)
Trustworthiness (Tie / No Pref. %)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Usefulness (UniDrive Win %)
Usefulness (Baseline Win %)
Usefulness (Tie / No Pref. %)
Accuracy (UniDrive Win %)
Accuracy (Baseline Win %)
Accuracy (Tie / No Pref. %)
Trustworthiness (UniDrive Win %)
Trustworthiness (Baseline Win %)
Trustworthiness (Tie / No Pref. %)
UniDrive
2026.06
74.7
-
-
76.8
-
-
82
-
-
Video-LLAMA
2026.06
-
20
-
-
18.8
-
-
12.9
-
Tie / No Preference
2026.06
-
-
5.3
-
-
4.4
-
-
5.1
Feedback
Search any
task
Search any
task