# Goal B3 Benchmark Cross-Seed Summary

- verdict: **GOAL_B3_BENCHMARK_CROSS_SEED_PASS**
- runs: 3
- seeds: 801, 811, 821
- locked examples total: 11736
- target locked total: 1536

| op | runs | total locked | mean routed | mean lift | min lift | max false-fire | min pair-exact fired |
|---|---:|---:|---:|---:|---:|---:|---:|
| `div_remainder` | 3 | 3384 | 0.909 | 0.693 | 0.680 | 0.000 | 0.975 |
| `gcd` | 3 | 2484 | 0.922 | 0.594 | 0.570 | 0.000 | 0.983 |
| `lcm` | 3 | 2484 | 0.969 | 0.966 | 0.953 | 0.000 | 1.000 |
| `mul` | 3 | 3384 | 0.865 | 0.852 | 0.836 | 0.000 | 0.965 |

This is an aggregation artifact only. It does not rerun models and is only as strong as the listed input runs.