FuryBench Testing Framework

home toIntegralPlentyChessObsidianCaissaQuanticadeSerendipitySpaghetSchoenemannClarityIlluminaPawnocchioVineAnura

threat-inputs vs master diff

ELO    6.67 +- 4.88 (95%)
SPRT   10.0+0.10s Threads=1 Hash=16MB
LLR    1.06 (-2.94, 2.94) [0.00, 2.00]
GAMES  N: 5574 W: 1429 L: 1322 D: 2823
PENTA  [27, 641, 1340, 756, 23]
NN: Faster threat delta generation The bitboards a threat lookup needs (occupancy, victim masks per attacker type, and the piece code of every square as four bit planes) are computed once per position. Records are written through uint8_t, which may alias the position, so reading it directly reloaded its bitboards after every record. The changed squares come from the planes. Sliders are looked up only through squares whose occupancy changed: a square occupied in both positions stops the ray in both. speedup.py, MinGW bmi2 with eval-ti-2-120B-permuted embedded, depth 12, 20 rounds: 1,835,108 vs 1,814,096 nps, +1.173% [+0.640%, +1.706%]. No functional change. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
LLR History
+3.39
0.00
-3.39
-2.94+2.94
0
2,787
5,574 games
AuthorWitek
Creation1795029847
Dev Sourcehttps://github.com/Witek902/Caissa
Dev Shac5d8d7727b0a2842efc66f161b1555c13cd4509b
Dev Branchthreat-inputs
Dev Bench6386874
Dev Networkeval-ti-2-120B-permuted
Dev OptionsThreads=1 Hash=16
Dev Time10.0+0.10
Base Sourcehttps://github.com/Witek902/Caissa
Base Sha0ec07379943a3fc1723a276150c537d5be12d3d6
Base Branchmaster
Base Bench6262536
Base Network
Base OptionsThreads=1 Hash=16
Base Time10.0+0.10
Opening BookUHO_Lichess_4852_v1.epd
Upload PGNsFALSE
Syzygy WDLDISABLED
Syzygy ADJ.OPTIONAL
Win ADJ.movecount=3 score=800
Draw ADJ.movenumber=40 movecount=8 score=3
ScalingBASE 1000000 NPS
#UserGamesPentaPairsTimeCrash