Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
STEM Reasoning on SuperGPQA (Pass@1)
Loading...
32
Pass@1 Accuracy
TAC
14.32
18.91
23.5
28.09
Jun 23, 2026
Pass@1 Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Pass@1 Accuracy
TAC
Backbone=Qwen3-4B-Base...
2026.06
32
Random
Backbone=Qwen3-4B-Base...
2026.06
30.5
SEC
Backbone=Qwen3-4B-Base...
2026.06
29.9
TAC
Backbone=Qwen3-1.7B
2026.06
22.1
Random
Backbone=Qwen3-1.7B
2026.06
21.9
SEC
Backbone=Qwen3-1.7B
2026.06
21.9
M2O
Backbone=Qwen3-1.7B
2026.06
21.7
M2O
Backbone=Llama3.2-3B
2026.06
21.5
TAC
Backbone=Llama3.2-3B
2026.06
21.3
Base
Backbone=Llama3.2-3B
2026.06
21
Random
Backbone=Llama3.2-3B
2026.06
18.8
Base
Backbone=Qwen3-1.7B
2026.06
18.5
SEC
Backbone=Llama3.2-3B
2026.06
18
TAC
Backbone=Qwen3-0.6B-Ba...
2026.06
16.9
Random
Backbone=Qwen3-0.6B-Ba...
2026.06
15.6
SEC
Backbone=Qwen3-0.6B-Ba...
2026.06
15
Feedback
Search any
task
Search any
task