smoke_accuracy · dvbench

Smoke test: fraction of parity labels predicted correctly.

Participate

measurement = experiment.measure(checkpoint)

@measurement.evaluate(params=0, batch=1)
def evaluate(params, batch):
    ...

@measurement.aggregate
def aggregate(outputs):
    ...

score = measurement.run(checkpoint, batches)
measurement.submit("smoke_accuracy", score)

# Next steps? Getting started docs

Results

6 runs

Result trend


kzheng/zkdv_probe_muon_2026_09_04	authenticated	0.35	2026-09-04 13:50:14
kzheng/dvbench_smoke_2026_09_03b	authenticated	0.42	2026-09-03 18:14:39
kzheng/dvbench_smoke_2026_09_03	self reported	0.42	2026-09-03 18:13:38
kzheng/dvbench_smoke_2026_09_02e	untrusted	0.42	2026-09-03 01:45:51
kzheng/dvbench_smoke_2026_09_02b	self reported	0.42	2026-09-03 01:42:56
kzheng/dvbench_smoke_2026_09_02	self reported	0.42	2026-09-03 01:26:21