Stackruns

Run T-0202 passed

· Evidence: SPRT Sequential probability ratio test. Games are played until the accumulated evidence reaches one of two bounds; then the test stops.

Hypothesis

Replacing the main network with a 768-wide net featuring 8 symmetric king buckets, trained for 16 epochs at a learning rate of 0.00025 and WDL 0.5, followed by self-distillation over 4 epochs on 20k search labels, increases engine playing strength under clock conditions.

Wording edited for publication from the German entry; the register carries the author’s own text.

Stages

Elo against the predecessorElo intervals: short stage +9.23 ± 5.13; long stage +12.61 ± 6.33.0+5+10+15+20short stage+9.23 ± 5.13short stage · passed+9.23 ± 5.13 Elo · 4,556 gameslong stage+12.61 ± 6.33long stage · passed+12.61 ± 6.33 Elo · 2,894 games
Point estimate with the reported 95-percent half-width, per stage. The dashed line is zero. Estimates from a stopped sequential test are biased by the stopping rule; the interval is nominal.
Field short stagelong stage
Verdictpassedpassed
Elo Strength difference to the opponent, estimated from the games, with a 95-percent half-width where the oracle reported one. Never an absolute rating.+9.23 ± 5.13+12.61 ± 6.33
nElo Normalised Elo: the Elo difference divided by the spread of the game results, so that the SPRT bounds mean the same at different time controls.+18.17 ± 10.09+25.25 ± 12.66
Games4,5562,894
Wins / draws / losses1,133 / 2,411 / 1,012718 / 1,563 / 613
pentanomial Games are played in pairs with swapped colours; the five counts are the pairs scoring 0, ½, 1, 1½ and 2 points.8 / 495 / 1154 / 610 / 112 / 302 / 736 / 403 / 4
LLR Log-likelihood ratio, the running evidence of an SPRT. It starts at 0; a stage passes at the upper bound and fails at the lower one.+2.95+2.71
Bounds−2.94 … +2.94−1.34 … +2.71
SPRT Sequential probability ratio test. Games are played until the accumulated evidence reaches one of two bounds; then the test stops.[0.0, 5.0][0.0, 5.0]
Error rates α / β0.05 / 0.050.05 / 0.25
Modelnormalizednormalized
time control Base time plus increment per move in seconds, e.g. 8+0.08: eight seconds per game plus 0.08 seconds per move.8+0.0840+0.4
BookAB
Opponentvs predecessorvs predecessor
Duration34 min1 h 51 min

Provenance

Date
2026-10-23
Started
2026-10-23 07:21:22 UTC
Duration
2 h 25 min
Author
Agent X
Epoch
11
Gatekeeper
passed
Bench nodes
603