tyrfing-12ksn vs master @ 8.0+0.08

Configuration

Author	Nocturn9x
Variant	standard
Creation	1793048911
Dev Source	https://github.com/nocturn9x/heimdall
Dev SHA	f4d1e67fce1d99329a0182c51949e91d508cf4b1
Dev Branch	tyrfing-12ksn
Dev Bench	5611448
Dev Avg NPS	—
Dev Network	remiel
Dev Options	Threads=1 Hash=16
Dev Time	8.0+0.08
Base Source	https://github.com/nocturn9x/heimdall
Base SHA	a010d9ec3b285981577b83d78a68c2f4a0e705da
Base Branch	master
Base Bench	5551733
Base Avg NPS	—
Base Network	tyrfing
Base Options	Threads=1 Hash=16
Base Time	8.0+0.08
Opening Book	UHO_Lichess_4852_v1.epd
Upload PGNs	FALSE
Syzygy WDL	DISABLED
Syzygy ADJ.	OPTIONAL
Win ADJ.	movecount=8 score=800
Draw ADJ.	movenumber=32 movecount=10 score=0
Scaling	BASE 1000000 NPS

tyrfing-12ksn vs master

Approve Restart Delete

Name  | tyrfing-12ksn
Elo   | 8.89 +- 4.22 (95%)
SPRT  | 8.0+0.08s Threads=1 Hash=16MB
LLR   | 2.93 (-2.25, 2.89) [0.00, 3.00]
Games | N: 8286 W: 2265 L: 2053 D: 3968
Penta | [65, 935, 1952, 1105, 86]

LLR history

Info

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>

Priority

Throughput

Workload size

#	User	Games	Wins	Draws	Losses	LL	LD	DD / WL	DW	WW	Pairs	Time	Crash

Errors

Time	Worker	User	Error	Logs
No errors