Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multi-table Reasoning on MMQA 3-table
Loading...
40.88
Accuracy
CPO
18.0936
24.0093
29.925
35.8407
Jun 3, 2026
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
CPO
Model=Qwen3-14B
2026.06
40.88
CPO
Model=Llama-3.1-8B
2026.06
35.19
CPO
Model=Mistral-8B
2026.06
34.32
Trace SFT
Model=Llama-3.1-8B
2026.06
28.84
Q&A SFT
Model=Qwen3-14B
2026.06
28.53
Trace SFT
Model=Qwen3-14B
2026.06
27
Trace SFT
Model=Mistral-8B
2026.06
24.85
Q&A SFT
Model=Mistral-8B
2026.06
21.47
Q&A SFT
Model=Llama-3.1-8B
2026.06
18.97
Feedback
Search any
task
Search any
task