Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Commonsense Question Answering on CMSQA
Loading...
72.89
Accuracy
Kite
36.1156
45.6628
55.21
64.7572
Sep 19, 2025
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
Kite
Backbone=Llama 3.2–3B
2025.09
72.89
Dense
Backbone=Llama 3.2–3B
2025.09
72.23
CEIL
Backbone=Llama 3.2–3B
2025.09
71.65
Kite
Backbone=Qwen 2.5–1.5B
2025.09
71.25
BM25
Backbone=Llama 3.2–3B
2025.09
70.92
Dense
Backbone=Qwen 2.5–1.5B
2025.09
70.18
CEIL
Backbone=Qwen 2.5–1.5B
2025.09
70.05
BM25
Backbone=Qwen 2.5–1.5B
2025.09
70.02
Random
Backbone=Llama 3.2–3B
2025.09
67.22
Random
Backbone=Qwen 2.5–1.5B
2025.09
65.9
Kite
Backbone=GPT-Neo 2.7B
2025.09
46.6
Dense
Backbone=GPT-Neo 2.7B
2025.09
43.98
BM25
Backbone=GPT-Neo 2.7B
2025.09
42.91
Random
Backbone=GPT-Neo 2.7B
2025.09
42.25
CEIL
Backbone=GPT-Neo 2.7B
2025.09
37.53
Feedback
Search any
task
Search any
task
