RefundPolicyAccuracy
App\Evals\RefundPolicyAccuracy
demo Middling and steady at around 86%. Plenty of real suites live here, and the useful question is whether it is moving.
87.1%
▲ 0.9 vs baseline
Latest score
· 36 samples
across 12 rows
· 1 week ago
3 runs
―― baseline 86.2%
now
Pass rate
83.0%
Cost, last run
$0.0214
Baseline
CJ690BSY
Aug 2
Test cases
The questions this suite asks, and what a good answer contains.
Runs
| Run | Score | vs baseline | Environment | Cost | Duration |
|---|---|---|---|---|---|
|
CBTKC9XN
Aug 16, 10:02
641ffc7 · main
|
87.1%
36 samples
|
+0.9 | ci | $0.0214 | 4m 0s |
|
VWAM7EDT
Aug 9, 10:02
bd0eb75 · fix/refund-copy
|
85.8%
36 samples
|
-0.4 | ci | $0.0214 | 4m 0s |
|
CJ690BSY
★ baseline
Aug 2, 10:02
b06e709 · main
|
86.2%
36 samples
|
reference | ci | $0.0214 | 4m 0s |