Qwen3-8B · scored

draft Qwen3-0.6B k=4

reference draft judged submitted 10-23 07:59 UTC

Passed every gate.

SpecScore0.547geometric mean of the three speedups
solo0.38× one lane; passes 0.38, 0.41, 0.36
team0.64× team lanes busy; passes 0.65, 0.61, 0.64
busy0.68× KV cache nearly full; passes 0.65, 0.68, 0.68
CheckMeasuredLimit
device memory161.3 GiB≤ 165.2 GiB pass
fixed texts' log-probabilities3.3e-07≤ 0.002 pass
sampled replies' likelihood+0.0164 ± 0.0857within ±0.05 at 95% pass
share of first choices-0.0018 ± 0.0159within ±0.01 at 95% pass
greedy agreement1.000none on this task n/a
reply lengths–1 ± 0.05 at 95% pass

Speculation

WorkloadTokens / verify stepSpeedup
solo3.49 0.38
team3.34 0.64
busy3.43 0.68

Speculator: Qwen/Qwen3-0.6B@c1899de
Engine: uploaded with the submission

Everything the judge recorded
{
 "ratios": {
  "solo": {
   "passes": [
    0.3802192971564448,
    0.41481622763384507,
    0.36329899403959226
   ],
   "ratio": 0.3802192971564448
  },
  "team": {
   "passes": [
    0.6481326567106688,
    0.6083271753368258,
    0.6351734342881076
   ],
   "ratio": 0.6351734342881076
  },
  "busy": {
   "passes": [
    0.6469640791759814,
    0.6798143862091586,
    0.6782119801033116
   ],
   "ratio": 0.6782119801033116
  }
 },
 "specscore": 0.5471385466303842,
 "greedy_agreement": 1.0,
 "likelihood": {
  "gap": 0.016351939678655003,
  "se": 0.043739938911346934,
  "candidate": -0.7005876745864736,
  "stock": -0.7169396142651286
 },
 "argmax": {
  "gap": -0.0018286344201578764,
  "se": 0.008107084788240597,
  "candidate": 0.15247568354139543,
  "stock": 0.1543043179615533
 },
 "fidelity": 3.265787269074265e-07,
 "lengths": {
  "solo": {
   "ratio": 0.765066056359416,
   "low": 0.4382676708360356,
   "high": 1.0918644418827965
  },
  "team": {
   "ratio": 1.0011997946930378,
   "low": 0.8566853505236907,
   "high": 1.1457142388623849
  },
  "busy": {
   "ratio": 0.9906627181146661,
   "low": 0.8708211180448283,
   "high": 1.1105043181845038
  }
 },
 "memory": {
  "ok": true,
  "peak_mib": 165184.0,
  "allowed_mib": 169119.1
 },
 "acceptance": {
  "solo": 3.4918080465341736,
  "team": 3.3371722682864555,
  "busy": 3.433437263794407
 },
 "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": 0.5471385466303842
}