Prewalk organic routing cost
felan-prewalk vs felan-no-prewalk
Cost · minimize · Ratio of reduced sums · lower is better
+66.0%primary outcome · Ratio of reduced sums·+43.5% to +71.6%Case range (macro mean)
Costprimary · minimize · Ratio of reduced sums+66.0%Aggregate values: baseline $8.5242; candidate $2.8997 · case mean +57.6%; Case range (macro mean) +43.5% to +71.6%
Metrics
Arm values remain macro means across cases. The objective headline uses the ratio of reduced sums after per-case median trial reduction.
| Metric | felan-no-prewalkbaseline · 6/6 runs · 6/6 passed | felan-prewalkcandidate · 6/6 runs · 6/6 passed | Delta |
|---|---|---|---|
| Cost | $4.2621 | $1.4498 | -$2.8123 |
| quality.passRate | 1 | 1 | 0 |
| Duration | 927586.5000 | 621258 | -306328.5000 |
| Prompt tokens | 4,974,393 tokens | 3,116,612 tokens | -1,857,780 tokens |
| Uncached input tokens | 130,834 tokens | 235,727 tokens | +104,893 tokens |
| Cache-read input tokens | 4,850,048 tokens | 2,879,040 tokens | -1,971,008 tokens |
| Cache-write input tokens | 0 tokens | 0 tokens | 0 tokens |
| Output tokens | 40,034 tokens | 20,541 tokens | -19,493 tokens |
| Requests | 53.5000 | 38 | -15.5000 |
Tests
TestBaseline (Median trials; Ratio of reduced sums)Candidate (Median trials; Ratio of reduced sums)vs baseline
memory-summary-links$6.81423/3 passed$1.93363/3 passed+71.6%
Attempt 1$5.2806Passed$2.1565Passed+59.2%
Attempt 2$6.8142Passed$1.9336Passed+71.6%
Attempt 3$8.3449Passed$1.4453Passed+82.7%
prewalk-checkout$1.71013/3 passed$0.96613/3 passed+43.5%
Attempt 1$1.6814Passed$0.9661Passed+42.5%
Attempt 2$1.7209Passed$0.9667Passed+43.8%
Attempt 3$1.7101Passed$0.8387Passed+51.0%