eng-0091
window recorded
citable URL: https://halobench.com/records/eng-0091/ — this address never moves; the anchor /records/#eng-0091 keeps resolving
65.21 Whmean 185.1 W − 10.1 W idle floor (box-idle-all-empty) = 175 W attributable
window 2026-08-15T12:40:58Z → 2026-08-15T13:02:06Z · provenance recorded
method wall-meter · aihydra energy sensor via Home Assistant (counter-difference on sensor.hardware_ai_hydra_energy) · no contention
covers run-0281
EOS-cliff cause-isolation matrix, extended-volume confirmation of cell B (run-0281, cfg-0076, 45 requests, 0 degenerate) — a cliff-detector guard run, not a reward-scored capability run; wh_per_task/pence_per_correct_answer left null. Wh/request = 65.206/45 = 1.45, close to eng-0088's 15-request cell B rate (22.012/15 = 1.47) — power draw stayed flat across 3x the session volume, consistent with the clean result not being a transient. Counter samples: start 19.3990371674299176 kWh (13:40:59 local, 1s after window open), end 19.4642435759305876 kWh (14:02:09 local, 3s after window close). Retroactive join computed 2026-08-15.
Wh, never joules — read as a cumulative-counter difference across the window; delta over the idle baseline is the only figure that means anything