Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Medical Long-form Generation on MedB-4 (test)
Loading...
76
SM Score
APCD
5.384
23.717
42.05
60.383
May 10, 2026
SM Score
FM Score
Updated 22d ago
Evaluation Results
Method
Method
Links
SM Score
FM Score
APCD
Backbone=II-Medical-8B...
2026.05
76
78.3
CS
Backbone=II-Medical-8B...
2026.05
75.2
78.1
Sample (top k + top p)
Backbone=II-Medical-8B...
2026.05
70.8
77.6
DoLa (low)
Backbone=II-Medical-8B...
2026.05
70.1
74
II-Medical-8B-1706
Backbone=II-Medical-8B...
2026.05
68.4
75.6
Beam Search
Backbone=II-Medical-8B...
2026.05
67.2
72.7
DBS
Backbone=II-Medical-8B...
2026.05
62
75
DoLa (high)
Backbone=II-Medical-8B...
2026.05
8.1
8.1
Feedback
Search any
task
Search any
task