Tests

Finished 25 shown

Author	Engine	Identifier	Code	Time control	Results
nocturn9x	Heimdall	tyrfing-12ksn #9043 Passed	diff ↗	8.0+0.08	LLR: 2.93 (-2.25, 2.89) [0.00, 3.00] Games: 8286 W: 2265 L: 2053 D: 3968 Ptnml(0-2): 65, 935, 1952, 1105, 86
Merge master and select A8607157 (bench 5611448) Bring tyrfing-12ksn up to date with master and select the new A8607157 network with the unchanged multilayer TI architecture. Keep the weights local in networks/files/tyrfing-12ksn-v2.bin; network publication remains deferred until the net is merged. Calibrate on all 7,153,653 positions in lichess-big3-resolved.book. At scale 305, getStats reports absolute means of 927.90 for master's 510EEBF8 network and 1004.55 for A8607157. getScale recommends 281.727639, rounded to 282. At scale 282, the new absolute mean is 928.76. Update the documented default scale. Validation: make dev and the runtime-check build pass. Scalar and AVX2 NNUE tests each pass 17,731 incremental/full-refresh comparisons with identical checksums. Python regressions pass (99 tests, 5 skipped) with Git signing disabled for temporary test commits. The repository commit hook also passes make test. The optimized depth-13 bench is 5611448 nodes. Network SHA-256: a8607157fbb8ac900ac7fe25d45d772509db9ecf3d9b5cb5000e5a581e1d0890 Co-authored-by: GPT 6 <codex@openai.com>
dan	Hobbes	bugfix/se-early-tt-overwrite #9040 Passed	diff ↗	8.0+0.08	LLR: 2.90 (-2.25, 2.89) [-4.00, 0.00] Games: 20200 W: 4829 L: 4760 D: 10611 Ptnml(0-2): 68, 2319, 5232, 2438, 43
dan	Hobbes	bugfix/nmp-verif-null-board #9037 Finished	diff ↗	40.0+0.40	LLR: 1.28 (-2.25, 2.89) [-4.00, 0.00] Games: 5576 W: 1330 L: 1292 D: 2954 Ptnml(0-2): 5, 604, 1528, 650, 1
Ellie	WakWak	main #9033 Finished Duck960	diff ↗	N=5000	Generated 2123334/5000000 Games Elo: -0.17 +- 0.39 (95%) [N=5000000] Games: 2123334 W: 757743 L: 758800 D: 606791
Ellie	WakWak	duck-count-stm-fix #9034 Finished Duck960	diff ↗	8.0+0.08	LLR: 1.08 (-2.25, 2.89) [-4.20, 0.00] Games: 15644 W: 5035 L: 5038 D: 5571 Ptnml(0-2): 358, 1501, 4080, 1552, 331
fix stm indexing in duck count history
nocturn9x	Heimdall	master #8911 Finished Error	diff ↗	N=12000	Generated 4312694/5000000 Games Elo: 0.24 +- 0.27 (95%) [N=5000000] Games: 4312694 W: 1434990 L: 1431976 D: 1445728
Ellie	WakWak	anadama_3 #9036 Failed Duck960	diff ↗	8.0+0.08	LLR: -2.27 (-2.25, 2.89) [0.00, 4.20] Games: 4630 W: 1543 L: 1645 D: 1442 Ptnml(0-2): 142, 458, 1189, 412, 114
Ellie	WakWak	adaptive-lmr #9026 Finished Duck960	diff ↗	8.0+0.08	LLR: 1.34 (-2.25, 2.89) [0.00, 4.20] Games: 16914 W: 5652 L: 5521 D: 5741 Ptnml(0-2): 379, 1604, 4377, 1701, 396
interesting adaptive lmr idea
Ellie	WakWak	tt-bounds-for-qs-eval #9032 Failed Duck960	diff ↗	8.0+0.08	LLR: -2.29 (-2.25, 2.89) [0.00, 4.20] Games: 9718 W: 3137 L: 3218 D: 3363 Ptnml(0-2): 255, 941, 2505, 946, 212
tt bounds eval stand pat
Ellie	WakWak	quiet-duck-malus #9031 Failed Duck960	diff ↗	8.0+0.08	LLR: -2.26 (-2.25, 2.89) [0.00, 4.20] Games: 9958 W: 3217 L: 3296 D: 3445 Ptnml(0-2): 261, 972, 2551, 975, 220
dan	Hobbes	feature/cont-quiethist #9029 Failed	diff ↗	8.0+0.08	LLR: -2.28 (-2.25, 2.89) [0.00, 3.00] Games: 17862 W: 4253 L: 4348 D: 9261 Ptnml(0-2): 51, 2164, 4593, 2075, 48
MasterStruct	Eunoia	feat/iir #9024 Finished	diff ↗	8.0+0.08	LLR: 0.50 (-2.25, 2.89) [0.00, 5.00] Games: 4608 W: 1343 L: 1296 D: 1969 Ptnml(0-2): 134, 545, 905, 580, 140
Add Internal Iterative Reductions
Ellie	WakWak	main #9003 Passed Duck960	diff ↗	N=5000	Generated 5000284/5000000 Games Elo: 0.08 +- 0.26 (95%) [N=5000000] Games: 5000284 W: 1786274 L: 1785133 D: 1428877
Ellie	WakWak	anadama_2 #9027 Failed Duck960	diff ↗	8.0+0.08	LLR: -2.26 (-2.25, 2.89) [0.00, 4.20] Games: 5150 W: 1764 L: 1865 D: 1521 Ptnml(0-2): 167, 497, 1316, 460, 135
Ellie	WakWak	no-mate-pollution #9025 Failed Duck960	diff ↗	8.0+0.08	LLR: -2.29 (-2.25, 2.89) [0.00, 4.20] Games: 7626 W: 2442 L: 2531 D: 2653 Ptnml(0-2): 195, 761, 1962, 728, 167
no mate pollution
Ellie	WakWak	quiet-duck-malus #9028 Finished Error Duck960	diff ↗	8.0+0.08	LLR: 0.00 (-2.25, 2.89) [0.00, 4.20] Games: 0 W: 0 L: 0 D: 0 Ptnml(0-2): 0, 0, 0, 0, 0
individual quiet duck malus
dan	Hobbes	feature/re-corr-bounds #9020 Passed	diff ↗	8.0+0.08	LLR: 2.97 (-2.25, 2.89) [-4.00, 0.00] Games: 14816 W: 3562 L: 3473 D: 7781 Ptnml(0-2): 51, 1710, 3799, 1795, 53
MasterStruct	Eunoia	feat/nmp-has-nonpawns #9019 Finished	diff ↗	8.0+0.08	LLR: 1.89 (-2.25, 2.89) [0.00, 5.00] Games: 8230 W: 2422 L: 2304 D: 3504 Ptnml(0-2): 171, 813, 2045, 899, 187
Skip NMP for pawn endgames
MasterStruct	Eunoia	feat/razoring #9018 Finished	diff ↗	8.0+0.08	LLR: -1.20 (-2.25, 2.89) [0.00, 5.00] Games: 4746 W: 1328 L: 1361 D: 2057 Ptnml(0-2): 136, 567, 984, 566, 120
Fix razoring in check
MasterStruct	Eunoia	feat/nmp-margin #9015 Finished	diff ↗	40.0+0.40	LLR: -0.81 (-2.25, 2.89) [0.00, 5.00] Games: 1584 W: 438 L: 468 D: 678 Ptnml(0-2): 54, 177, 338, 191, 32
Increase NMP margin
MasterStruct	Eunoia	feat/nmp-verification #9014 Passed	diff ↗	8.0+0.08	LLR: 2.91 (-2.25, 2.89) [-5.00, 0.00] Games: 9030 W: 2677 L: 2593 D: 3760 Ptnml(0-2): 175, 928, 2239, 984, 189
Add NMP verification search
dan	Hobbes	feature/re-corr #9004 Passed	diff ↗	8.0+0.08	LLR: 2.91 (-2.25, 2.89) [0.00, 3.00] Games: 55324 W: 13300 L: 13009 D: 29015 Ptnml(0-2): 125, 6401, 14334, 6662, 140
MasterStruct	Eunoia	feat/nmp-margin #9013 Failed	diff ↗	8.0+0.08	LLR: -2.26 (-2.25, 2.89) [0.00, 5.00] Games: 6742 W: 1915 L: 1990 D: 2837 Ptnml(0-2): 212, 821, 1363, 780, 195
Increase NMP margin
sp00ph	WakWak	direct-check-mp #9016 Failed Duck960	diff ↗	8.0+0.08	LLR: -2.36 (-2.25, 2.89) [0.00, 4.20] Games: 6928 W: 2204 L: 2297 D: 2427 Ptnml(0-2): 163, 685, 1839, 636, 141
sp00ph	icarus	cuckoo #9011 Finished	diff ↗	8.0+0.08	LLR: -2.09 (-2.94, 2.94) [0.00, 5.00] Games: 5246 W: 1020 L: 1061 D: 3165 Ptnml(0-2): 1, 398, 1865, 359, 0
Revert "switch to "standard" repetition detection" This reverts commit 4d89ed12a236770dbd7894e220a63809ff483afd.