Benchmark

SAGE: Science Answer Grading & Evaluation

Compare models to automatically grade written responses to science questions

Tasks
Grading science/math
Grading writing
Data types
Tabular
Eleni Koureas on Unsplash, generated using Gemini

Leaderboard

Rank Participants Model
Best public
Weighted F1
(in descending order)
Classification rate
(in descending order)
#1 0.5673 0.5724
#2 0.5664 0.5544
#3 0.5657 0.5000
#4 0.5548 0.5421
#5 0.5545 0.5632
#6 0.5477 0.5551
#7 0.5418 0.5241
#8 0.5272 0.5099
#9 0.5197 0.5312
#10 0.5169 0.5258
#11 0.4966 0.5130
#12 0.4956 0.4912
#13 0.4939 0.5078
#14 0.4925 0.5040
#15 0.4872 0.5220
#16 0.4861 0.4981
#17 0.4630 0.5078
#18 0.4619 0.4562