sakana/fugu-ultra
Sakana
Consistency
How often it gives the same answer when asked again. Across 42 questions. How it's measured →
Alignment
How often its answer is the one most other models gave. Across 43 questions. How it's measured →
Hedging
How often it qualifies its answer or declines to give one. How it's measured →
Answers most like
Answers least like
Measured across the questions both models were asked, at least 20 of them. How it's measured →
Where this model stands apart
6 of the 43 scored questions where its answer wasn't the one most models gave.
15 runs · consistency 77.8
15 runs · consistency 77.8
20 runs · consistency 27.8
15 runs · consistency 77.8
10 runs · consistency 58.3
15 runs · consistency 40
25 runs · consistency 26.7
20 runs · consistency 40
30 runs · consistency 40
10 runs · consistency 61.1
10 runs · consistency 100
30 runs · consistency 20
10 runs · consistency 40
10 runs · consistency 26.7
25 runs · consistency 25
40 runs · consistency 8.3
15 runs · consistency 56.1
15 runs · consistency 50
10 runs · consistency 100
5 runs
20 runs · consistency not scored
20 runs · consistency 44.4
20 runs · consistency 100
20 runs · consistency 66.7
20 runs · consistency 40
5 runs
10 runs · consistency 100
10 runs · consistency 100
10 runs · consistency 46.7
10 runs · consistency 100
10 runs · consistency 40
10 runs · consistency 46.7
10 runs · consistency 100
10 runs
10 runs