Qwen3-8B · scored
stock engine (A/A)
Passed every gate.
SpecScore0.997geometric mean of the three speedups
solo1.00×
one lane; passes 1.01, 1.00, 1.00
team1.00×
team lanes busy; passes 0.95, 1.03, 1.00
busy0.99×
KV cache nearly full; passes 0.99, 0.99, 1.03
Lossless gate
what each check is →| Check | Measured | Limit | |
|---|---|---|---|
| device memory | 161.9 GiB | ≤ 165.2 GiB | pass |
| fixed texts' log-probabilities | 0 | ≤ 0.002 | pass |
| sampled replies' likelihood | -0.0182 ± 0.0842 | within ±0.05 at 95% | pass |
| share of first choices | +0.0044 ± 0.0155 | within ±0.01 at 95% | pass |
| greedy agreement | 1.000 | none on this task | n/a |
| reply lengths | – | 1 ± 0.05 at 95% | pass |
Speculation
| Workload | Tokens / verify step | Speedup |
|---|---|---|
| solo | – | 1.00 |
| team | – | 1.00 |
| busy | – | 0.99 |
Speculator: none
Engine: uploaded with the submission
Everything the judge recorded
{
"ratios": {
"solo": {
"passes": [
1.0109584631074073,
0.9995725467752457,
1.0026613802593292
],
"ratio": 1.0026613802593292
},
"team": {
"passes": [
0.9466879411583641,
1.0336010034686982,
0.9957247514934774
],
"ratio": 0.9957247514934774
},
"busy": {
"passes": [
0.9937867908277217,
0.9891039953482672,
1.0310447743241948
],
"ratio": 0.9937867908277217
}
},
"specscore": 0.9973837084928465,
"greedy_agreement": 1.0,
"likelihood": {
"gap": -0.0181673517427573,
"se": 0.04296733878786788,
"candidate": -0.6590672417342399,
"stock": -0.6408998899914826
},
"argmax": {
"gap": 0.004370347437437749,
"se": 0.007894688924244617,
"candidate": 0.1425569643607589,
"stock": 0.13818661692332115
},
"fidelity": 0.0,
"lengths": {
"solo": {
"ratio": 1.023233702401053,
"low": 0.7957246897090475,
"high": 1.2507427150930581
},
"team": {
"ratio": 1.0090914762991863,
"low": 0.864911563484329,
"high": 1.1532713891140436
},
"busy": {
"ratio": 0.9973155773536605,
"low": 0.8909749129613697,
"high": 1.1036562417459512
}
},
"memory": {
"ok": true,
"peak_mib": 165782.0,
"allowed_mib": 169119.1
},
"acceptance": {},
"limits": {
"greedy_agreement_floor": null,
"likelihood_bound": 0.05,
"argmax_bound": 0.01,
"fidelity_ceiling": 0.002,
"length_margin": 0.05,
"memory_slack_mib": 4096
},
"reason": null,
"score": 0.9973837084928465
}