diff --git a/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/README.md b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/README.md
new file mode 100644
index 0000000000..155dcd6547
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/README.md
@@ -0,0 +1,50 @@
+# GDN-Hybrid + Sliding Window Attention + compressed-code warmdown1000 (cold-cache 3-seed mean 1.01671233 BPB)
+
+Joshua-owned SAFE_SUBMISSION confirmation run for the repaired GDN-Hybrid family, reconciled from pulled TensorPool artifacts for `run051-safe031` / `j-bunuyb7fxu`.
+
+## Headline result
+
+- **SAFE_SUBMISSION authority:** `quantized_bpb`
+- **3-seed mean:** **1.01671233 BPB**
+- **3-seed std:** **0.00134386 BPB**
+- **Best seed:** **1.015700 BPB** (`seed 1337`)
+- **Worst seed:** `1.018237 BPB` (`seed 2024`)
+- **Artifact size range:** `15,713,422` to `15,903,365` bytes
+- **Legality lane:** **SAFE_SUBMISSION** — fixed-predictor / no-TTT Track-A; all pulled artifacts stayed below the `16,000,000` byte cap
+
+## Per-seed authoritative results
+
+| Seed | Steps | EMA BPB | Quantized BPB | XSA BPB | Artifact bytes |
+|------|------:|--------:|--------------:|--------:|---------------:|
+| 42 | 2227 | 1.007164 | 1.016200 | 1.021202 | 15,733,879 |
+| 1337 | 2242 | 1.007164 | 1.015700 | 1.020105 | 15,903,365 |
+| 2024 | 2227 | 1.009032 | 1.018237 | 1.024111 | 15,713,422 |
+| **Mean** | — | **1.007787** | **1.01671233** | **1.021806** | **15,783,555.33** |
+| **Std (sample)** | — | — | **0.00134386** | — | — |
+
+## Why this matters
+
+- Improves Joshua's prior staged 3-seed SAFE_SUBMISSION artifact `run039-safe019` (**1.01710033 BPB**) by **0.00038800 BPB** while also reducing the worst-case artifact size from **15,981,262** to **15,903,365** bytes.
+- Confirms that the compressed-code warmdown1000 repair generalizes cleanly across a fresh 3-seed cold-cache confirmation run rather than only a single hard-seed probe.
+- Preserves the strongest Joshua-owned clean-lane architecture family with materially better mean BPB than the current upstream Joshua PR #1564.
+
+## Technique stack
+
+1. **SP1024 tokenizer** with a GDN-hybrid backbone (`[GDN×5] → SWA → [GDN×5] → SWA_shared`).
+2. **Fixed-predictor / no-TTT Track-A lane** — no eval-time or pre-quant adaptation in the scored artifact.
+3. **MuonEq-R + AdamW** training mix, EMA `0.997`, late QAT threshold `0.15`, and **warmdown=1000**.
+4. **GPTQ int6 + zstd-22** packaging.
+5. **Compressed-code record packaging** for `train_gpt.py`, `architectures.py`, and `configs.py`, which recovered artifact-size headroom without changing the trained model family.
+6. **Sliding-window attention side path** present in-model, but submission authority remains the pulled `quantized_bpb` values above.
+
+## Legality notes
+
+This record is **SAFE_SUBMISSION** because the scored artifact is a fixed int6 model with **no TTT, no SLOT, no RLS, and no eval-time adaptation**. The pulled logs show all three serialized artifacts below the 16 MB cap. XSA telemetry is reported for completeness, but the submission authority for this lane remains `quantized_bpb` from the pulled artifacts.
+
+## Provenance
+
+- **TensorPool job:** `j-bunuyb7fxu`
+- **Canonical bundle:** `~/parameter-golf-project/jobs/run051-safe031/`
+- **Pulled artifacts:** `~/parameter-golf-project/state/tp-pulls/run051-safe031/artifacts/`
+- **Parent repair probe:** `run050-safe030` / `j-br75brdy6j`
+- **Upstream source PR:** `openai/parameter-golf#1545`
diff --git a/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/architectures.py b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/architectures.py
new file mode 100644
index 0000000000..0191a703e5
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/architectures.py
@@ -0,0 +1,5 @@
+import base64, lzma, linecache
+globals()['__file__'] = 'architectures.py'
+_src = lzma.decompress(base64.b85decode(b'{Wp48S^xk9=GL@E0stWa8~^|S5YJf5;A7bs%Uu8wXbs4O{*J2g^P4X}8A*?nKeo}D<QO>((oAtmlzXSM1Ig2`L`+cv@?w-@3r2bpI{8yz9`;@SXxb8{$fRtA3$`M_7l-P&$A)VpsI<{M1(k?)Sp%fgGHf{OA(e1{%%h>U(XyoMXIPp+maZ<`Y6aE!)~96}-?&KS_k=U~h?SsP>3(#hSXx;Gb}nY(jTA3=m<)>eOpgmMpv(N6Jw4N1{;sYZhPMhU0&O&Da!-;hJL0RkdivHzjmv^(5uZsfYeq27zGmIe`KoOHIQuYUjFMPp8l;u48XK7Wszs%jwMD3Ag^3BFZA|PV)yGfF&9VH-(s2Q|eiU=*2gP#b(g1_keckp+RTd1ebYAlkI_Q(1KdgIzo^2Ajexh6Vo<-BmT64<v4-1-B@Zcy2g-N}82Q<Ea*=hw3AZ>)g?{FY-8_?)s$@TwD&jnwEE94uXTfL)k!^9qGM3)Q`34S(SfW9B%*GWL@QNjM=A7Q9M@<u(gRT%zP@#PE!V{C+UBo+1IHl|$&CKbB<9~V0pkE~35`x<jBHXC}G)3sBisMOXu$^%bGFSi`<f~Bd4%sn={P=e{86L<@A<n--1wB}LOJz25ZlC_=nzccg{Rl}=S)3M6&X1NoN>9S#*t9C)TqR=33=ksZTjaD+X4M-pW80`pt#r_W8U_;|k+Ua{=p6)56RyNUVjjaM=XPL{;3+Y<$iDw6ztH0YE_r!|+t|eW=X<2P0w{@zd5iBFY$CTrD*n+ol+(E42zm}zhsa#B?qtO=*Lc{nP&hfDdkRtQ=Ep3rLq>hVos*{y!iUw=w2kYqLrMkW!EgT-IHvop$I@64dqxYqCV@_2s5AjSTf;6;^g}%+S<b`Qlt)i$@u^$(%=<Fn^$~BO+_|0snc?HA~$-Ti$=^!Tfm~Jpf1xjpCNVbO#!Oq?1Dt<LPzo-LQgJ(5gaj6CQwPbR{*wHUXA6$Oyt`%y#MgUM#!UXz)%o#5@p~zjNN;Jrc;n2+g=%K2-n=5zHKFi4zka>Zf;6p2bFm~5gVdaZ;l6u*JAf0kul^?N2uz0!sG3i*0K+|_D7}U1uB^#T(Yj%9S7V277Q}3e^KEP|TnWsSNT!n98yW)qh?F`Dne{<Yd1xy3SY#uYTFg`Gj+SGdkCCBPoH0p&1O@jA7plGXa)Mq+MjT+dxJqr!g(!@l`OV>&h-HQW_;jl)yQ(JD0+=SXB6lPhIl5y&+;)mb@J7se?c1nHe2f;xjtWiwr%87?ToT05m1D7Ht^7wfyHwC&oUu|2#kt3@McM!NN*Xc(8l3poPj324}xWH7HqS;NT18Ws9RV~4Jb>@o!@qB$3Rod_D-2^PR;B@BW*;pxUq=9KFG^`q&c=d>!=yylYsps!c<pZN}!NhMnN{~GJIrND%t-DTGz)wKBgH4AJY*7d^FyqK~@{a~puQzF;@vMm_nCih6o=v8~gI3otWAl%rWCqUSJ6);n)Ef9N0C|DI4tWZc^)Lw7eHWzO#k_VR3VF}2DvvASmgC<&p`7N=oYe#$d$yl{ynjYr4-5HF`DW8nA-h4Oj1GLGQGH5=dV($~R5)q;33CtFcM1>DqUp&i52VmLj6d^p4h8b^NT6U4z9CbVh$=ki+|kz+>sEkqyPVAkwo(WKMju~!K33A^GmP{d5wpd}cC<c0BF>>rGEH^_v|sZC>pk-|TC&?x5C&1Rrf_c^uk4hePX(VYWmHYq2yxenyjs>Yfo1y@K2Xm;pgPofy-WD7x^L`xCmBLInjBy|B5rtRG7`-X_RQx_vA!=wCi5)p_NE$)u#5le=Kip=y)gOL{>W!g584gaFlfPn2Mm`OuckQXPoC`4ykJQ?t?T~;IPnr@c6(Y0>Xi(9;rSI3(mK55h2yOusT&))QSD}Xr$z@0jT(LP!Z8jfo5%LAlprOO7;9m>d9})Qqc99#5Y$8$n!Oca<Hw2wMR8E-=mr`fW{T_@y8%GS<#1XT8NzM0$g~1tg{9rg%C){FzX(U6X#auj+D$l5jV0Ss-4@%z;)(@gYv6`-<G}F>g9SJ|J(&AkT9tR3=Ep8tYdo=BP$nEgDSOrG2*Uq%bHKitAx#+Ox%d78EPqpwjakDR+Dr!tnkELJ>vP9y;{XLy;%eyP9P@bC*dDiovwN|2OvUL5|ICP)ik0h;e=q<7iKiW2(+=ipjVsA%Pm5lk8}(T+hMK!GrqWo6mWTqJjgnh;h8&w3Z6jO=ORx!i`I#8ku6W*#@azgGJjEFo@ZeoRuq$0W!%8xJ?=Kr=H~3nLT`CZF(}?vt+0&L@w3j`2UpupFq;As;(s%VL1G}Y!UFNn#ueXiv06o}s21CoicTpwuuVeOz75vlEsW%_<QuA+H8P?GACiUkRgMGO|+Ww<*>(K#<Lx3aF=6eqf;3gQ}<ArSb{ItG|J=C!^+q*t@z@9N%AIb?)XEh~26$_$w=Ez+X+!)7r?|0(W11JJQbYgp{wX;2p`1+<a6PiG4Ls3vFit1qyRwLR>f5hlGL#P1P0#0h|z(EZidY3Z8dMji`PH&efpmyR-o?}(jL`3gmDksLIN$l}HL_7S*h~v{e6xsM3a_zc&5~tYyPn|q0(A{fJx}h4x{?J+=Kzj`#?i5q!K6NZy1OIDyReIZr^o8JPPlV6hY@fJB8vbE<uqRaaJsmhpP6|#JOX%v4D2(Wf(shGg?MvQ=6d~&e*{G<8y|(1M#g!t5n>&JuS72MPoTMbu{uoW)YXTJy_`|NXcDT1mgP%}3oj$nc?vPe-m|cPH5DvMkj=T`32U1H?b&!Cai5c@^>mFTtU(q;h_Srv2GN+>N-!`S-qe`6yhU`4v&-`Zf1vmg6DOleRZ~=7PiYtq0xqkDAs1^CmavP3W=1rYCO^60AD17J<L1>NN*DO@eqx3EQ`meU=>bQ+}A^PKmF3)1d@zn1FqVE*3t2MZ>lDV(JO&LEMhl)<l32)#M#1RNJ)C>?7qmOO%lI&Byxq`w__IVlA+oaI4yAP~CY1%&f`4TTAxcT6#bGA?~utOF<)!z7&Uc5J8o&=;`I0_c%&O7lxa`;KyQBtc=v;6(+C~@~I-0h)Pklxj2{t+K*0b^vtA%r|VB(kekV4P=(+)bGWe5QIezx}!<l1b0*-wzK1=FCT~!2tVA@ymFeM5?7X$C&)cI{1Mc`Zb5kmI5bN3kYEo;2)W!yPdA$zuRWE3KcS=sx%~u!%F@ng!{-FQ`QD(4mA$qsus5B9=cGp{;b6Yw~%2JW{ByNvPmG2-5)2OY?mEW;J8xTI@joSgNEPnFu80XHXG3+Ruqz&wO5iL=j}1g?$C0lHVBSXcj(wl!9C$_wKO+d__}*yp%o7CYJkx=!`H&iX54R9WZ<Pp{FPkFtQ{BB=F=-yd)C5%4CxJ|)#?7LEAKjzcSYqhpT~wSzVS9Gwc@%~8J^h~iK5ZxkBw+Tn?h@zscii#g?$D1g<rN<s%5;fFA9L#Vmh1eK5sdGEI0F;P(?V$u@x2lr}QxTI~SB~q8uP4S@W>&X}#)nCr|F>@$j+G<rRiFM|u!;tBkA>n&w86a8m4=i_-z``GsN9()Fg#o2F|P*D)ivEjUV>Ii^x)CY)P!er`IvAsr1PSZrqUwq*?>JMP-aa(VP?Y`}EA0{pD_VYM>t&J4lqU#?>MC-ZE94FVL<UIn3rr>#)WRp}ym5u$zku}HFYK^f|e8*?;gE$fQ$g-=xai0%n#7!{L^7gyoh_RQ-uSD6Zr`O8?{`Bs>&WZEo2&+eRHw2LpjTu+#ef^s_DBpj&NUua7j=EgZH(bR945C$%@ZZLQ^c-gF0bMI_BtQxdZC>LFk-5F^;S3~Lq6MdhzPLYzlUvzGWJJKMFeq1Ne7!$^F+-(Fpeyhelf_G6bHf8;ZGVidY_ko-k8M@de+KHwM3|h})k)Y>0rzcDk%{k?+2Om?kA`nFUp5ydxZ3HUfwvftwoBG}<oty@%e%lva9h!-%pz4e;1g(h(1msuu;VxLq21iM*9eHDHWkAo6wpPCxRb6yGBp;MR#KYxw(s?JC^$@df3cEI%5y=cBh731+Zir^v?DT}*Ve(F{RKg$9f=^Xsz5TZ_%4hrhDvlx6v24c{#4`CLpuRALavL;Xoh&XZ+5Rl6TdpM4y8ONI!~Rv^)2^r0VWdSdnv1_-;$3pdnB$p2{hUs70TaP>Dk0$%IJEGyMm@zZm~kF|5PF;9`zDGw7gLuzP82|o0yyo`ZdH>sL3u2P&o7}?C2C^PZ3myM-DPK)MS^7ikg0XYSiiNc1DMuW8$WCb*7Iy^>jLH$!uP3v5*g*eN=L2B^7pl9@53$H&cEV5i`ev*&$3ZCZi2<gGa1Th!a8>zhED2vUfz<p_qol0K$%a1DJ%Ltvte)4^}D2LG(7IWip+DHZ1{U-*2uNm&Z(6jcAq|R&VSIAFLrdY%kCdG5EMYlwN+{yZ9JQ$1{P!~n>7f`VwL+_`2+UD0LAy^9r;IcJp|FF<R(;JEM9+`-S=ldW>H2_YRp42GchCW1&r>Llk7as4FQh#y~L;lp20-;&-ZSzR*4Nt@MU~O5hK(+@9rQ?Ofq`;i)w<&D;(t-%agZq-U3vMviP}g<pm#*oHTqULUUq60Xhy71;cYy%X!e0C9@Ud*Srx!i3;5)-FbC3vmCq6g(AeH`DByozcER*b`$1&ed7Qh>Icwv0ry_`#uSdQa()%F&Jy%ptAx}+@iu)78qNMfI(?JEO*f5)O-n>(VW&5IjzA*lWs(S>6uZ6TNOYKN11|YqS)(ljRR94GHeG+uJ#ipZp3!F3nI9S@&0Gxu3d(YkjaPn#O4sJUC4@~eG{3e%kMierjR2g_`htusm|mB5)Y?)NRB|>XYVr$4od!rGOhzb6HXk$Fe4-jnr=?AU#HNpv8lqu0KpeomI8)I!PU*_}Eqo1&q3Em@EqlI>MPuI$R7<?B5j)1u*yJ?aQlc<$DJ7386vN6rkVjfE=VMcK=8M9J88EqzsMBTUS&gUOhh+96&b9;_Wqv8&IzL4<<~fAPz>Q+D2pCybW{OtorvR;iBdUlx1MkRW{}FHsjWaS*(-GE973A#%V9?Yc7$fNmwpIzh7;7cbO3eVQNitYRK{}WiK3SPBZ>PRZP?i>OvK#oCGXWcm_&w?4xFI9>P-&+H*XbbRJ9k3r)&YxqN9*_OI=3JS9JgRy*CZ5C&7>Ews3w-?#Ea@1iy}1bEZ`+4vmPSXPoc|lejLHIoU%sL-w$ygnC_j6<<n)-wnN{R!gD=nK2btfMoQ$d8{(35!I;*)i#>DjfUvISBKCOYpjiYgd&995_C?GN6Tj=MXumdE5pq;`PK^FiVa9Q2b?>*e-MtQdhH-xZA`VjHtAZ?Pt@h!(YS3dvQu5{RF_8?i(*qKG;qz0H;1LWnQMd=^6x|yTDQ~|X0DkYu9DJfZO@AR&Gb^Y$o-t{50rkVm!Y@uz>f8EqSmB2np~O?XMG_+ZMfKa2SY@gBv(FynMU>}>%1uFDlhkXvHMlNP$*x+iS_;Xve#O;2d3aS3wPO29oe08K^Ykwt&1H=ERDD0`tc2sQf&m+B=_=DO|4m7TX+9lC96R2FE-L*0fnx?-Yf)>xbdtyp@o?8WsT!Jq@#D#s8Qr2U;pU>~21-#ZvNj+h7D1Sbqp!S`=%xcJ=Os-b;Pew)*Cp>J2-16oU!O;DR43+)p~a=k>+o>%p-CFfl}-RVA75zTo`{rqu@y{M_E*GbenUK0kRh@Kp%<-9-y60jZ5HOD25_4GD4VJ`%Rr~TIb1M@8^lLdE`8p1LCW-LH}ZV(B==Tsh`MNvfIM`{$NU?47L$t*QjoP2>LCt+d#1qTPVCCClPGMc9Xf?xWeG?64q(0tV^SEQX<vB)twN4sVp}B3P{OOV<Qgc>yRdKp@(>@x3=2n)HVdb*;ie%yr1H!y%vxF9jaP^8RUz9eCG@_b9{?rKWpQ{<W)Qze^=XHO2S;IX0E?6?T)>BL3?J0l`29|+<>~K*M1f6(vNaQN8rt?%u$T7BSZX)^TNcDPvRves2A%<fY`@JjTqAb8Q#$fUQw1Q~nb|c6ptd9E`dE-QUXtXz_4XAY86>%MpG0(Sr=Gv4j5_t50%ibaVnFY{*L5~!ZHYS#-}c<Rx_IR1k@c#GX6c@)Am*R0Z7K8Wa;r!Zr;k67S=93_w~)qx6-X4KB@94!R!rDSwyl%#=w~=a8bqlgYbKh3_9YZ4SOCLj=nLyz1}~czhtDE<b;mEPO%)vqvYa=Rs#qFmKY~)gfyttDcc`~=hW&&S8-8PRgq|qg_cf1jLFFyG`Cw&6-$V}yKg1qTgvW;l{&;jCL`W&I<rmAw_|ocF#UVmgba=UNn`z@QlJuOU&2&meSjdnCgsm$S6tDMhue29@p-J>9N7Jyv2+%rQU|QbiWnQodW6%sie}?-af1QRSSL2lC8mVgVj3|}%9j2SWE-UJA>GNU0AN2Fw3ifb(bL8d{T!^))d^Y)H5AkbE(EMiY$)x&GQ6Fpisw<+OKXb9`jj;>`P#l3#%Rb}g<ClJ0`%(0vTpQkj;)RMW9O}f7fLRP7Qz#9iHeKu@$$WYbTQes(pGET4_D`kiGl$6Tq*dc!&~-axVfq^XlX~s(I1;4ArdA*Pg`C4px^mH_Z%M~r;tNG2RS?jLqzq5G;Gku<0+h*^&K+QKh{Di<XU1@{+5_S4Q%PCRIEF1^`RIQPlj?D%-of1g&+Mp&(FwA<k6bbM@hR^RI7L#|<Q@gTSQ_5lp9*QbX?DE41d|X6F!KfybTpa(0}au__NJ5l8B{><u9SwridQDmraT+}YMQ6A-5y?q4AV}!FShrPOXHo9_kZYeN+-Zyt?w|G#JLyzJgCc0*jz~4x!zysE+F|cUo0i6XH#{eIfK`V5mnOFx%=DRT$%aV6?m=OTjSp1Wsjqy{0(gZV1%rJW!K-)dFa9rBEY#=y|P2(ZGILJ8vn!u08qkpR3{2O$T;n<LENL`^9Gs3{o`CUj%8QKx4DhWy^N3s9RhwQgUo9O>Lk21)TkQrqFCfO7)__6_EQiI+LeqisY={74y)gr+Z0OWyXJ|O`3v*~^^#A3&HNYKi2_#&l)GuH(e$VkJqE~bRpE6D8pj)1)Pa<JTBMg<49e`0RNtp@Jhj9$fWdaC2FJRx_GKiEg+W86emX=yb*~#V)xwXnFU)#SB4;8zA$}aUzuIlw!#;)NfR}v+)j7EKF?~`(G#x-wfrF=Ko=pObG!AJ3=B$1do+>3IA)${Bb9W;KBJ)pDVVL7{cdGE*T>C_PE}ZNHWrUbarV+~pL=7j!>NZxVqBVT(ICrMygQ-N8G;v*LAj0e>5Mhjq#(Z`Z;&fxC_9TQz=x<$7AOogW^`A7jJW;_r5IqZ$Bb|>@v~B&7&NGAQJ4O<;?uCxMSV=qU-Ism)5zjl`o^=yLf>(64MGO^z9M~$?N|q;%{j1wxL+6ctZ)TK40ELOxc^V|EGyEa=CqXK9IFaHo0cR&X)e<Lu1m36Wl;56UD7o=bt`6Cjq0r7rx&eo_wr2%+tVfSa{`{2<hhw0~_X4{<ScaO3v$EHM#UdJ$z}cRS{bkrMg}6sb*j3$9`Q*N|bGanaGu~-E1O==`R%5%EW3R^lFaeSl#pSuR`~b>QI+?Osg~f9j(P7w=>`hhB2oLhAwba|b^l$6Zw{BB3D%pICv~{&e^*@?yX%`kvM*}lFXjXF`QXN`V%%Lbi5gECrFIu$^j*rjGdh6c33!Lnx0H156gF$<(UCSY8hnwRqTu+?Q(Yq2{RF_@!|Cfhz`6nO321;32A40s#PAQid&SP&s=*eQ2cuA}g8ZFv;cP)S)yKG-wFkwU#LP614f$)#|>(^5-@fr-vlPHx&J@S}O8Q&gSdvrEOD?~m-oDbtouDTi?zxg||zS{xX#D@l3fs{L+IaFLTS7OEswe)16cOv!C0S-=R0c-51+&1Q;i7mT}4?&S3bFa(M8X-rgpf63<&Qa%5?XSs@RXSBPYB&La_icC`VZDlfF>rP?=<BU%sR&g4w(M&zgu`*!y|00<j)_f{7)~rc+}aJ!0s_R>TAq|bXq!vl_}3^Sdt_RTm|e4Jqe*@p^8H*fh-p$FYdl5_a%6$faS5bcqrL3XLP{O#VFkV&)A6g~_cTjOAOGor+Rs^v0Wpui4Aav$W_SPq!EW3W;+CKj00HMQ+Q$I^sD?+7vBYQl0ssI200dcD')).decode()
+linecache.cache['architectures.py'] = (len(_src), None, _src.splitlines(True), 'architectures.py')
+exec(compile(_src, 'architectures.py', 'exec'))
diff --git a/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/configs.py b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/configs.py
new file mode 100644
index 0000000000..91d93596ea
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/configs.py
@@ -0,0 +1,5 @@
+import base64, lzma, linecache
+globals()['__file__'] = 'configs.py'
+_src = lzma.decompress(base64.b85decode(b'{Wp48S^xk9=GL@E0stWa8~^|S5YJf5;1K==%Uu8wXb(;1LpQ-~@JH;7bq<!2rM>sfvNj2i-6-~M*R0t23rwIc7?+`^!VD6uagcoK6DbS{9T~420xAWik!$biWb)rrE+C-SGfkiwhV0~r&>A7MzvaxD&2QbJ{2wIdK)(s%&mtZ-@C^j&Il;>(HV9Kdzl26iSpQ@}+W8F+gFd4uyfywfyMFM`tZ6yPk2@}Bf`F{`G|QqY_~SLO)^$j(lVw5|@~26gr#@Cy$?g#DNu)(MZ3zM3D848h-pG^4;H!1seE8;!)*!vr<*hh_M)n_J*FkAv3&y+~EaG*xEwZ(y0{Dluk1)M?vK|<?M15&9sh9@)sm=9EQ!l9<F=3hI{V_#FZb(VhKFRJTCvj$idas@TZXBNDLU;Z*-1PC|^dt;z;sGN1$)2AK@%4k~y7Ij08HFB(E_tT8(BIU{1VYBaM8e3);f(C<O!9w?`>r-z=2q;$f?Np>RAmtd0SnvDX9u|NdpKRv1XCu)-hRaPhu(+3vIVgK=%?W`h<nrj@^ClP57G3H$Ka_p0^Lhhy){F)Rzt1mZ4=&&wEnHFK4UVYgugE;)jv{i-?&ldyTBS(mqu&f5X^|fs%L$&s&;0h<_JOdpj5J4B-WfP{dfC9+{^sg^v5;>*Ir!o`XM-gyZHpY(;A=WQ%uIy-LEpk?32}<@uwZHb{=xO>@YgZg&n)w-2CzKB!f!PX4{sbfc`i@s;=Z)Js>=-iM6SYI+Q(`U`M<IVh8XMFDd+B`sqIQM5;%)xR;J+D72uYc+Iz<mDBMb+;NowsU><Q-aeFf6J}m)nODlO?Mv84+2(r)&&4T&86TFCk=yXWPs~%Vy-LioO;Yd>gV{}<wkto7{`3JW|8UjLBhjgROZzy&cg`aRWW%}9i{#DRsIpaD9_Va<3?@d{W`OB1T5`_6m`oCMDbS3M7BzqM<WQX(Nvd|X!atM(ztWo!P$;vfz$#P<5xGfqqg7;Yr-so}n=-sgRPRsFA9Fl0F%Pb@36LC1gv#~HAKY1Ml^jurbyw~$C$RHK^VlPtr!MLXq3wTKOCFZ_P3-79YWYpaCz_BWcE_YNptPThG8=ZVmwW2lGk<RbZgtH|@~c!AI7~_|56dJYhXhj_=hL<YLZz{c3mGo}L36^tMzj=fJx_u*k6!31n!ByTnf?<Me-7elw4W;X!EBx6Cs~oZ(X}TztV4Z82EEqhG0{8x{QBoG9kwpH*|JMc$m8jCmaz7f5-J%nY(HgqSIzgomvEiNbsB~$Tk@6G8mHWLI=%U?eAy?aPo6mX!x#wD74cjcK>Z-mMs{`4u_n2wsGC14_Pi&1O3GbJ82-<=%ePMG6(KXaL$5W57RMv}0_vMs`@UkDBTH-R%ko})Yg?AOCht_=5RD#cvY2)jQ&tE*`7?1>pJTzY%sLiRkLAqYJr0qyGdhd~uy@0D5xG=3C?waVq~$|Q#!<?>JCX&-oh{sufJ>q}>z-_UbAeQ3l;2VRB$MCD-_NXhRZ#7v%p&MFs!t`_Y$`GiU4-YOmqjOJm$ZOflTGRdW}mWmDGGLnOW%_=;rSv3Gm$hgn<~9@hs)#-g{3+d;Vo~}ax{{l%l=VDe2Bo^B8Lcu<75bM-hIE)J8?E}g2~HGoyh5%fYjL46`UGK$i-9Ta^wXSK-7bJn%)z#XM+i(MFA#)^J?TaE?@kIZ=V4zfp$|^j3BuWKRuaz*OmfluO>S6^XY2?e9Z3q2+1CVv*C!s#1PtU@V=DVxEFsS>8lv-tc~<X(o7~cC62?Guo8WU2W#K{m_c%|lavV8mW;Tmc+d*b>n;l5Q+cxVMb98Ou7vHP5Xf53+_dNGNVPzcF-5<Y@s`%l@sGT&_6tnrf2ac(0YTLqkxs8D|5jppc#LkSCX%PAd`v&<7r1YYXy}Hg@^(PC)Kz<^CrSVSBK&Om4$c`m00HL<{~-VX(Bw{7vBYQl0ssI200dcD')).decode()
+linecache.cache['configs.py'] = (len(_src), None, _src.splitlines(True), 'configs.py')
+exec(compile(_src, 'configs.py', 'exec'))
diff --git a/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/requirements.txt b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/requirements.txt
new file mode 100644
index 0000000000..9068ce0f70
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/requirements.txt
@@ -0,0 +1,5 @@
+flash-linear-attention
+zstandard
+sentencepiece
+# flash_attn_3 is pre-installed in runpod/parameter-golf:latest
+# torch, numpy, and other standard deps are pre-installed in the image
diff --git a/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/submission.json b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/submission.json
new file mode 100644
index 0000000000..565287603b
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/submission.json
@@ -0,0 +1,43 @@
+{
+  "author": "Joshua Martinez",
+  "github_id": "joshkmartinez",
+  "name": "GDN-Hybrid: Gated DeltaNet + Sliding Window Attention (compressed-code warmdown1000 cold-cache confirmation)",
+  "blurb": "Joshua-owned SAFE_SUBMISSION cold-cache confirmation of the repaired GDN-Hybrid family. Pulled TensorPool artifacts for run051-safe031 / j-bunuyb7fxu show a 3-seed mean quantized_bpb of 1.01671233 (std 0.00134386) with all artifact sizes under 16,000,000 bytes.",
+  "date": "2026-04-12T22:30:00Z",
+  "baseline_pr": 1545,
+  "lane": "SAFE_SUBMISSION",
+  "evaluation": "quantized_bpb",
+  "val_bpb": 1.0157,
+  "seed": 1337,
+  "three_seed_mean_bpb": 1.01671233,
+  "three_seed_std_bpb": 0.00134386,
+  "artifact_bytes_max": 15903365,
+  "artifact_bytes_mean": 15783555.333333334,
+  "seed_results": [
+    {
+      "seed": "42",
+      "ema_bpb": 1.007164,
+      "quantized_bpb": 1.0162,
+      "xsa_bpb": 1.021202,
+      "artifact_bytes": 15733879,
+      "steps": 2227
+    },
+    {
+      "seed": "1337",
+      "ema_bpb": 1.007164,
+      "quantized_bpb": 1.0157,
+      "xsa_bpb": 1.020105,
+      "artifact_bytes": 15903365,
+      "steps": 2242
+    },
+    {
+      "seed": "2024",
+      "ema_bpb": 1.009032,
+      "quantized_bpb": 1.018237,
+      "xsa_bpb": 1.024111,
+      "artifact_bytes": 15713422,
+      "steps": 2227
+    }
+  ],
+  "artifact_authority": "state/tp-pulls/run051-safe031/artifacts/train_seed42.log,state/tp-pulls/run051-safe031/artifacts/train_seed1337.log,state/tp-pulls/run051-safe031/artifacts/train_seed2024.log"
+}
diff --git a/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/train_gpt.py b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/train_gpt.py
new file mode 100644
index 0000000000..32d146623a
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/train_gpt.py
@@ -0,0 +1,5 @@
+import base64, lzma, linecache
+globals()['__file__'] = 'train_gpt.py'
+_src = lzma.decompress(base64.b85decode(b'{Wp48S^xk9=GL@E0stWa8~^|S5YJf5;J|A!vRwcXh(bq-BQwKy4>SE$m+hy=@aif<Hp(x6uSQXPb6c0-VX>6he-v&MpW=|DABH+ur<2v|dE{-umk+De@IB<?o<$#h;IvkJFed;l|LGgCvnN2zxX8<YOf>O_%R){)H38d%^cLt3<gY=jV==<uR!q3OrQE?#ia~#ZrhF0LejHG#Om1k#%|M|iz<^Tcqcy%9DOF?(TgGj)?5${`1!cXJKtfV4DEkQE>+fKpUUu>@n&H*74d?B``ayAS3A+#)LK`<;#TH9chg3g$Vev#d9;6rkcw2z{8Av5<wA^6&a<K{83T}J6is0hCp2zj}g<U9;D$ljC1E_y@pG%={HESVj0B{;o=_A^wV95IlP|-1+JiZ-Kk%O=?a9PGP^e5RH$F0ycTyh7lHN=?WIJ$4Js^i>$KuXw~HQp#qHmbZr0tCM0Rv?5(Z!T{1VE0a!<P*92C!=hmJ998Y+|B{{)Zj=`Hi4!WjvBu7+K%|z#i!HIP|}>W`HO!^WR%`g5t=J2qTdZMUaKTx;aydCH#sZ@<=&YBf>_JrF2h4Ui1DF2j!4AS-_r!0TXguJy_Cd2qb!UnH_w6VF@x+Cx0<&bhA%;63e0<*03>d^cN+aIA^A7t-(rcNfI6;Xu$2a{l7~qD*SgrOhYLRfWcDOMD5Lv#<!FB}Lg;p25qFrH2jTrYax;<GxCnD!TCJi<`6y9@WkbR|y~O)c2kEgCc0ssCx(o}uxeD}YwJh%U6jCbPOazLRZ|kNNwuGkZv1tw^1+=4r(=;-Wlez-Wf#?rZM*FVn(czWX4$3vToWWsTKp@qlPy=<zji>lTqr9A*Sm1$8q{^TlJtkb^$SY;&`s8wb$LhgsM8{Q<7+^oN0f3whU(6?O#{<$}x(6EOaJ6g#>7G(hrva1+Zw9DbJ9~=z2MxkEWgc(?Oh>q72Y%rmXZTXq^1PT3V+P_fOmj`5E)7W<+4YyPa?o`+LrqH(x$S+lc;?G=7TPayCZ_<Bd<|(pCRor}!+0WEp84oDQ1Mg{^>ObR+uX*RWr!B%cgKP&d9r$5$N#GFU4BYpJh4vL4`iOT-Liko>=)Ge0m<u(t!x()=)nuFTVmQZzz<H*O@_WAc1_1A`2MCnTu(#8<V!1ik*Pt2%TdalP{nqWHaikY+JYC_2Z%Y^!wul~5xiY%_l+9SrkB=+G`WX2Z@6h*+&^9@6Y45lLR-x*w@4ZrRiVeWoNm|uh2YKg(M;Y`yrT?QQ$}86C8q&JN#)$5(@9;9=0PhJS2YHPzbG<OngyWu_kR|pX1NA!nNQk!LLrx2@gL^L@88CS?&_7j$vxv>;hkVL>7N%;Jwg(0iFhj9)k>9aO3p|fXIHYl@|JtB-j>b=_+B>2rO|k^*?`Q;02H#PgMXrmwTLVdIviw&YKfXNQ_cUJ(AI8dSv^blDyR`|<beXD!l@$OG8eQ2&@O%!Uf+c7%9{x?F@2akTzw$PH9X#9`dQj>Dt8L8+|>6O-~QzD0uwl*?Kb662HO2G)deDN*-_i2Buq0Jj)^HB1So(hc&xx-C(;4^&S)^zdgo7Q+pqtgue8C|_~NXEFB1OlKW`KK7B9i}d$;^j?XjT$%Klgj2Avn{RKNWLXq3&n9w$2-ERv5$anN$N9&bmW4#gf?MSr{Pc5s#=#|N8$g3Mr8db3%e+hu7x$9%1<r#4YODDx<$yi3v>iB7JqbZ#B$Gs0wc7o)julo*-*XWOGEFu_v53ao$z5T1hP7WY$Jx|5}ZSds!FUZV2V6~eXkzs*4E#{{+1i1l01Y@N~#TB~gIn<H`)g#5;(Z;A7O9AhKzIl3s_&-w&5K5~C1oELQXPn9RaAGyNr60!$@$9{_~!ke?gT4rpaNEtW)qVR~YRBZ!l9fr@{d_u|*JU4!nmjj@O6n`!SaGYoW7h~Qdv!UnqzW1BLjU)Hr?q~g%73{pnGz?Qv1Q|B*VFmZtdH`w9wQb$H0?drXrWuvc1^0r(wqYyBu&YI$(v4MEHiW=bBPI{;)8%xCj#pw!S+wNLb*qEh=9&|@g}0M~fx(L8k^Mq+NMk9iK*fm>NVk|xG1F$rgVBObj!*ldM<*|;Ng&40bCo6AGx%#DXhP`_q6skPSu&3*eVN1feFJ%!El`iBz#>&8_SQ~IYrbWQ6{>;_&F<i~3{Ep-QT91C-DYcnm)j4~x*Pbyd9f(Q1cWh+kJT#q=${e|y-2{XW87xx7uE|<oZ~@!Ko~$1bcpfv9_{bO)=f)OmkoT{mYP1JggQ&NP_S+At2pr(nWRj^_omXIIKEbvpO9%$o0WAQ-@i%W4lrp%tUT?%pkcQ;4bu{`DfxW(`t1C(a?jpY1LRhlo@e>m5P|)v6^Ya?98&GM-4H-jIaz0C$2+7ayF4$#X4`GpNAF3n$2^*|(mdbX);jQo0ct1+!OKL(aClgAV~;MS4z3%E4T4gM6XbX>2~TW{tT{4pV?K^v<v2}z^!a=!)N(Tlv=i@(SX)=LJC>Qf@uIM!BQc=Nc_X|nyv~6XxDB=@ZMz{lVB{^(oQwytlsZxx7AGnp<QyRToyI(w*$gk%ZMm-5iqLdPPLiKDx}rD{@dV{|=hocue0QlH9A7F`b(;{aVX~$E3syZWq(<EByKhMZrVeC(q43A3A`*6wt|`0Mq2fJ5mQ&i?06~jDIlitc$U;&8L<hz@@PYVzNt8W)HSD?0gnBY07o_xgDx1b_FwtAL3rR=mn~QMPm>m8Y=xn^jYB`nH<{<}iH^YB}3nZTcajRqheKbTTLXijTQQz=fVf@=3R!VfRLBxeRz$_nnEZa+cj^zV?lTlaVj)Z*jO|}{8dA+0dfL?kT!^~J*sH)Gv8D(2fbAKt8J26|;?5eZg)D;YU0|H?^4>PJTatbmw;i7=vDrdydv+jEDJZ>MyJKsN{<<)A*?G=C)O%s^n-RkbSg8~#GMS^Pm!ScO#Tvm||LI!2PY?|G!Q1662hJDOgnE@xklT3D@Q&GT#pf-Inv)mj^podkVVsW018~WysAL~8T98Olo&qHR1zjnqs*RgaO+*q4v%*5#Ypkr9+pqujEBz;Hb&?XH44Gj$0)OqTHFhVe`Kl=-yu`G81_q^I7>aZw%<SYaYaUc?<eeCG0bxX_<_jL(xY#%WZWzbk*a!>aJ@{$TtegYfCi7uUV5|W_SrZC;GL))Py^zuf#{-t3e${`$~1mmW8?D%k7KUD?fA2}=-Q-@bB*H2$DoT9znJk6$F7?d_4YPA!^S`VO3iH!VT;xCbvDqF9bZ|uFftU4*kWSdiDxl*`9;C*!R#svA_^It`wXBb0@{w<r<0A{`=+*5*-mZG6Ao|KueRd-iV0Hlv@LSh=UBR?fm&5Kt%Eq$9f$=Oe8vEb>+Bl@cE>$*^|<L{pTmF5!L?p;z%Oq+Gt56!`R<3MfJ1rcGmQX%X0B~gA%OtkPnYY3#e8EMIxg}eyQj=BsaM~94(x8w=1e5VJ?1}WozL0{v6_l-;`(qe|DA}?yEjE8Y#uYVVSim%|oy))FSo2${Fk#MC|GOKk;odk$wSwLYUsP1-?f1KxxX&jJ=hCjNw8oq-`jW<kvI6Oo~WJtC_y9flnrCOSjFX*W*($$_Kgh7X!^S+~WbIQPbr?d)#5%)xoj*@mnMfj{6`U7RI!F2Hhw&#x%i0Jmzt!Qm?QPIVFsJdJtfp^miO-kxPTyu6oHFE!%xUOJ=tGB^ZW~6cV?HQwWJXa%ywuq4Re$r;HEry`BfB-Ob<Zx33_o55Ea}++$GNYa~gw1Am%XXO7Pz0b2fxR34W0)a7=;p~)`xM#s`_Txf<Cm%|*hRVtIh(fE5L<E*S~R>bU{$Wu-37z@Ub<ySVF59leDb&9sKhfk3n%?~9H(Dij=H(R;jdBvB#BKLYcsp>y&*Ci>heneX7FYuc~1$lYH^uQVO8PjS@zqDE!dycJC{6I!9^oMg6n1D3k~%ULHd-Ro&mt{tJCH)A*9ko)HsfVH}uAJ-40`TxRQ*%V#IYW$`>%1hv(F7l(8<(rnTGY)eBgNS~-F~+W`GnFL39C*TW9?%yesmSB*MD$MgAOMy45Pxs7%NKD2C%aLSRJhtE~&xVM8ZY^&4ONUvZuQ1|AG3JZ?#BY6e3EVJwdia!^<hvXCK5^T4=wQZg%4w;>k0Gn@Y;7Q{(R(7+DfeHOD9(&{p8WrMmBEl02hdB>2?|sFLw-a?8yXMXS90(#=R67x!b@jsv9ajQSHx>ao{2<GAHJc&QhjgdRWvapS|G{GS<_RJZ@6)*0Pe_Ywx1sTBqHgh*9Z)DL0Q0YGE}YyS$TU9#Fc!23uqB0aBU+I(zuEy2njjL?K$S?5S#G)diC3&VFspP$?<$^U{E%-}q<guFVZq9w`6Jl-eM4<R#HU?+JNsL_PS~@|>SQuoFiaQ;jw$D)b&Jh>lu1BWeoxTNu198z*-HLFh8)UBaP&q-0$)h2y%g^i1i5^i*7>oXk&*O%r*mahsT5Cyd<<Eaeh>=Qywf;>MnM)^amY;`lbr4FufDXb(g|0e55wznIgrk`&ahNV4!G&7EfH3-!*Nq&G5l8Qx#?@)#eM?yMes)*xodFmKk>(cy2a2)@#zoZt6t@9943p<FZK5umg52kgu;GgZC!)OS^m&yA>SrXUj5)`7U63hsAmU%YgXu2h}wTkPZ<q2LXGv@<!c@uLOeiA9Dd{<M`x>P+ng5MGMk8tAjX_8KWvmjLvK=`Lct8)1p9Z{`P+|f3|%VG(@RA7dDk^vyy)G6`T3t~A@R3#jDRg-&HeM%{lCIS?aP#<S;|CSsai{10~NAz%fxN=n@?Gq=!Cwmox)PvD0(Q_Cuq2u*0Sn<q=y6sz|wKaG-olxldqzc)Vh$@*Y6VmpwtaTQF@WbW6`DMM|O|JE+6Lp5m-n>5<ALmuTczSg1IEEaRi#%7eCsjO_RX67|rgXJ5jEv-V&h|^fN;vlRP0S6Fmg6SVh_G4cu}y=cR${^kwD(<?;&z4hY`v)Y-4i`s<+Vd-(oA+SzO`>&xn_K2Jsa-<71Q=h3sIjQy}K;67hjtJ2akcflfh#%Wf_c@D~P^OZ>^&<y%vI;X|}y1|c`A!@O!)R#@Od2ov~)s=m{?$3(41E%AGkXTz0(XIO@yG-N&+I*Qiqi(jn5P{U_H?oy{PPXMv5LNmJb1+X1I#!C<hZE_>merOlbC*&@lHAA~3XbmZ;l7CodYswrU4Y82cIAS7#l8`wt6d8OTPf;P1`8gEOaH7k_hr~Hfr%dUmMexf0TX8kgX%&=U74k*5{n1h4Uvj~x;!HM(p)+8AEvC2H>UI1dw?r<?Z_RoNmtjLkyFVQKC7)*$;(B^u$+}3_afemhbW$g={{-JJ|-%YKHkQ@Lh%M@C9RN2?HWkl>U;GC5=?V7stiq6OD>R#V>_yn0~;-2kRo%JTp)&|ajKWXWEeb;;akNf%xieP*q8}P)=)MXwu#jplNa&`TKAs$^D$=p(2!9u!dM}j`*3xeKPS?e<1hNMEJIT@uhvsw6HEP>^qG<*1rNRhc4;a{*@;d>!z0O+&JriRk3rBY`=XX0UN!s{>2rY}Ai?A&t!U0BYmlJQF%t?9^Nnx^s<u}mSWNZ7inXE?v-w4|!5Mm#$#E7oFYC>rc`{%%sUl;jWuSAsz<;Y7!Znkk#`qu)QP#L~<Grwq4`cW%rntM%8D&*Ia6I>WO9i?&-O=WI12gtfgO*Mhc{t(04dCI@JGewwWWCTHt_tp8P)MBQtxtJ3KQt3uPv^_BmAwy2qsae6fAFV7EhBcfVQp!aJ=Rjo%{D~9=-MqXf5<0m*I7ihgsU;6yGH?~UYX0&|6Bhj5!`?`k7M-w=)Cg+#&^)?_hi!rInT>)qpWPMj2VUFOk4KD`3rtr$lG|o`$87D^=)!a3~+jY$$s{Y!!T&6A=N&JoCHiOx#H5QP*Q_I$&hky@Qmc)L2Uln2ehCCrjY7#lR%%_>gNsO346pUVMdFcINKwthhbwP5s7Ke9P|_hcM6?xf1De>S=Es%qbTX}#hA@n4G75-ccN0)9v;1O(iBjyfJa-JdgS(0aU15Y5d;__Y=8bPk!q@<lk0XxZX7R{31wU_YbYr#x+7#*KaK*Z1xQf#fOT5Fy|+#5W{ioa6>d?bNzhiu7F>w+l_D8%THoiry08K0fK%)Y>Ktovre2wkDE$m7h&Ak2dqGaX$j(XC$?7zJA_{~cJ8s~&|MEhmy=UreKgJpIa`{c%Tw~`?FMi%hF1bfcnXvIGPA5fa)PEp+!lM+@u!v{1NHbCLi<B<fu%UJ3*y~Ie2YG^=3v#ii_+WYdVLA3eP^txr8rA1X^}K-CZx(_vFanLSXh597D=wxrI6z5ZwCch_+?d02#MP6k;@?@<Yt;WbNg<V9BnFW1bnr0sw)}_JTmmDIGFx3`Z=t@NhAC|6`$_96Y*I&5f3P27yj6u#{@YgY>Y`$XEA--eL#*E{LBu9_g`Zfg;;bo{_f6;a18*#Pp4+#If<Vthsxebz1;ywgQVLS0<ObWF-KGYiGkVmBjv=CZ)~Pf6iXwC_&1+~#|D5WLlHDfY-jYlj!24u3-Gy+LnnhG%?{SeitZgzf!g<&nGAf?ui`>AjdYOWg`=Z!Mzm9)(uk;j(hj}{j)li7|{mFfIueu)3hNq4^AWH6XNXKYg7<gqK3Zl9;)cC|F#$&`PtJWPhS9Hb_t6)~D3_&Mk@a+36MqEYuH7dS58-gU4b2ZbFk4De5`m6NyJEKYs+q$hJ3GjalQ<|cAbF&wtKqRVet}V$|JB*pvLB)moCJCJh#3)CPr8Tt5QaaND>%v<&AC=v@dSLPeWBZ*G3Z#IP7=T@*^4!e{{jUNGHTBYO&jk|uSEX@QB}4xx9mZB+{~=@-Gn~T}L+Wtqxp@V$wba^?d86Pl_$x5Z-(rp;f|A7s`z|3hIJ3)HMw-fy3qBOd5&FP09_vDj2xpT+nGQCSYGB(~O%N(vD*)i|iFo-)@!7~gT8PR<C55DKg?S~GW#(<fqx4Xv#4Ypd$pEQ}Fy3)p_DOoz1lygBtV|%|8u)Qy^T^nQVOp-rkQ1e)bu3Vb)f@|qvhIG~ef~N?<{kr4Rr$gYtW&T&>ti6$3T=7vJNt-m+!dsNx+GPeIT6^>HRlb=M=?@z{j#IYuaQ;n@1hpG+vx?7t`L~`Pabg1e;4OiX=Wc=<fsIjS*chMAW`Q?qjdNBc<T*GbiO22NBfo+lB3yqgLS=6=|spjY-tudB#xJnYXGq1o`&`Ve`~%(1M2wfg^YvPI+;A`a;(SxwCE`YmQn-S#@=-XbvA%sbr{d5+EE|!(6bQMN4p+{aYDvvoXY79$A@ztw&<M2!tf>kiE~A-SZ8DXs^rUNH2BI)Q;u*>mlW^ntH26bPcc|3g@>cm9`vY}I2^ePey&tk8T0vcE@ay_aAgg9E>o85oo(6;RY)m$&ry(aRkRC#!VNUGG)b1YZn-4E!;1-@^Dv)JGD#WUm&Mfq;bV5c<@`5j_2Vef#mi-LF>MUUrHCrIW?V|(Bx9Z9dW#`d>yi=U@BoAwlEo*`;(_nq#x$gI7b<tCy5dqz+*O5K>wwq7vEAJ4+GY8t)%-v7GTj?qLvQ)Lutt|!H)um6SYOu4*9v*h)neT%1txS3D5oEivS0L6XnT~ATN7Tm$JFr5tTK*8K5&KzWVbsn*c+mB6F7vs+xt}MiGZ||Zhe$#p6ZNCy_7(c%BhJo7Mqndc&tF?KQ>_98Bu6LkU>?$_x$eG%_K+%`cDkDBFc9hHK(wbHO>pqvQdv5xyemr#6QD@KUrA7t9{07ff8>GaX_uIwz(hrd-yIcws3Zed(~K8Re?xe;sz;F#UR_hw#=Aj42H7PU*cgduY<7>Zfuj7*{Pt#+Um<*C|DQM5r(Q2j1a_ec;NutN!Hb0T{x#^6JzHEt$SKl^^<nC{YW2<vO_A_NFzt~QZq;5C~9xFpz@T%>Ldkw;$T;S#w!uk%224sZ_h^dk-+`u3Sr?^5x)0pqmtDzH_T)L+jz{l-zNGEZ*B<UGq#gjWhR*Opf?5!1M%2}peO)J2HGibJyl!Z=4;fXc2I1ib#8!H5Q5JKGWR|UBJ|W3TiQVZsP>AYfa|7GDgz=Ql~9{X)fUFQo^Q*`2Om12L?RsQnYgY?0%*x6D0NjC2=|0eW8(ZSt3HBh%h?=?c7qw5CxDf%02$gH(`K{dksbKHu|!3PKh5@JIvKfa{BF0p&gli*#{V+m7<&hfG$%9vJ$fdl#2zdGjMhA@ZZD-H%zmid@~Cd85e|jK=U|x+@Oa=)Y+|@IUnlV`kjQ|N{uHQ#2Qi7riMwubnQ)pll5Q(c%WMhkOq4}d8`gN#F(@6~e@?jbYhv;J-XsXnTgKNP+R8(egWvt!Odt_<ggW`@xf&@cobZ&9Xj5EaUE(X?LdcIeh;(0Q9yMgaV{Qj6_WXa@bJ^H9@8i6lP15~_=rwC7Xzl?a|M&517qC63R1r)|%iLQ!Bl=12me!o2^J#p10j)ea2ZeZ!4H^mt+8j>=3&3W%VuBIl;}y9&DPY7`R6t9cwl1?$GRh=J*Ie8nxn{6RyJPKFSylHS)8H7p@fmx)eUA<(4R)~%)1vkIpSGd^tX7&s4nl-WUMxO%$%zPOF)UZZ$^bALQr#@qCKlQE_ctPDwy&hRmbeKiECOrnA*I~=79HvBnr8afyE6!`4>vAS(SCaRRDW|!)I%es-<BvwQlm`2n%>8BX77ey+<uyU(3lsDP8FaR&$QR5w7E50oGl7ojZhN_`yfgQKmfJP(6pU_Rg##fiL%<00_|BCa#kK-kJDpb1RE^J)noC_!5QjgKW9*-g|I_@ik5?A3gUmrshEV34&)wc!wUJzx5!Cgo|-p~!W^+#2+IsfwG;yZV$V01sz3KFV<HvfO?65HTg#kQE<gB#a8g{f*(ZuX#=gzvHgThnK2#FZxo{Z^BNFo3qwoHTFBbtKkEh>E2z1cz?*)(>>vSHRf1wacuBy(`qaA<}i0`;ajD?#S|Gf$_q})_Jm`}*LZ<_9h>0^vN(8qPpxO^e`_Hz=i&xWm9VDQj`FuN!c5M)i5_hShTLDzw@;u|nd7bxW%P#3n~V2(*tEtL7cCKzhScv%^+Y}qJ^O%U<nu(VE;wfe{RgK9L+{P{)HFT6{=eZTs_tW0L6&^gNao`}~ctGF;jpX5dRLl&0P09a1&HbYt1*={v9MszeBuaNhmZ0X$lBv#+@5uKV_yb1Y*22CXL>}De_Iy^vwcl2a5#W9&PUBY1{9EmMx!`rzyA7X+{#Mn-viD(F!&%vPNG_~GUdrhKYF3JZVD?gsOC#U_eLPJchljw44ag=zh5Iwn`eGitQ`zizx7Qw6J>AD@HNNX*&KBrMHnA=;9tW%O^S|6&)bW*Vz5a7&rE6CD!$;u`Wm;z%Rd*>5G7P;Dfg%Pf5?}6H*?Ia+TIyj8OJ?(wkD4H9pQ$3GOjD>qL!R$>?III1>NE+msOC5oQnL*6mp54{DWkXZ?gv1P1h)Ej6e27n5^#C-&)q@<76;LT$QII|;{BH6XOQ$PeY|$<ne|RUsX#oj44@jJtiFBN!R(d^g6yz26e+as1Rcm@w?o~o`2I^$qk937!3t(yrZJ}!mpfP(Db>ez;l)9@V5&A}RYMoiSLxk~=OB~rg?A;wLkUbgPni@1!{cR$qzJisDgu+pJ<KGg&tb=0w4T@^);2bkVIfSbOj_b|PC<d_)CK*vmvDuDO5W;du={M&`Vpz>q*@hIeXu={iy!tyzFuzkb`pe3uR;Z&9pSVwXl6VeqbGzV1@|V9~78@ua()PMU@#4vy&6<azC#9KGI<m^A$jt{wjtiY3|6+|m4)#?uR3*UL0?Q&xC&s&z_nb)V3^=?>3_e(?czDcTI3#n2E+=ta_1v&B2*zW0^v0oO({$;o#b{qh@k8ZQ7yt#rR&`4r<oA_-AzKypuO`CbU^zTOB(-B~wVQ5jNx06d45H(wK5pb#!xxsTgGPj1G~uGp1qv?tk;zQCb{FLw)>AVfhS%-Mxwr(K>g9z_khL99t0LeQ1_v&X{0X3%|Dyi0<@{SRbqzA5Q*@r%@|_(_6uzMOZq)3LH~ABH*X+7H-q!yYs{4oI{&qLM)`*+&e)EQU-_wGm&xcYNr%_`|W97VaSmAaB*1r`2Fv{NLOfky1;~S?t3d{2bgACToz`^LhfY7f7suUovD^qG|-7zt`*0Xv;_8Z}a!FIfKN~>r$#LJ9_h{(Ms*q|`*hbS+-xZ$0!iAjw6jhf%cva0-r-l%5^f;0%b%_@Mf0O}D5f#AZKTs?bD%5-C-jfTzI)&%D*Fzd#zOlBnf1Zn*)9TNyr{NY%gmTqcC?~>+`@Xwx(h~*<y9~xAyHk3WYElcB{ggKT;L-Qa|o$;x5%XWWN!>}zqD?53}-m`Afo+Q%2B4cKlMmY$}IKupA?Q2k_U7p?G<g#_o#vC+ZYyCRUm%vo5R8p8BSugh(418$Gvx0Fd)4Z-<s^m<cn}j(}W0O)&8R}*#_JyRS5uaIyIpaZf6$#+bb@nA9>Ot^mL)TjMF*~5@EmZPviRTIDzg+7tEMC+7kNh#uj<6h`zs^8@=Zw~WZEGZlY>T-Ab>H~ZFsj{C(<l1P#)yxmZ`g$g({$t+W2@dx)ZG^+=Dn5ctz3xfj`haag3hKO)sp$CzQ!T9w`!eqjZF3N@+tOH7?u%F*3R@x<1Wdw-(vEe8Nyz;XZ{E$p3@z1y%U})9jyT2WV=d?#XZFMgX_IKY{*jAA8*f0!?cuIBw|pVVrY&N@wyN%cwPI~<MhQkao%0=&L7rNAw8&rS5X(=2ICOI3?&)%8as}VnOBTszq}Zn>!^iS2<P<8)e(a!MJ3PdywDu5=1vK0B}Xsx!9t=8)32QDkzoTnidm594sFK9MMuGW<iQ17X7J_Lh$I#KpP#lY;kGLnj0JNC!_MGh2p9(djHyM4=yFpKke4$HVG|ou|JKtA;$)`r1F2NOVMgdO*=Vi@Ss8BeF#H4CLDMWkssRW(LA1gT9%@Q)+)bqJR-`)3X+sCtiRuWBd;kZmltD3EvEJd^C0#PrOLGdc?vmva_L%2eC_V=sdQ{mMV`5oUWCWWNElqmg#hlX_Pu&LxnxV}&6V2fnAZTqX)e~6?o^M=*i#uvFSYd^5o&-SVV!Kl^=$>*AWDVk_Cao*XkcVhbHeDcF1y$5&sWUR)*h<$9b#+~jTP!t4nnlFaSxmf8w{)sq*n%%))VrxI(AAe)1~IjWkLo+yb2FFX7NiWK3;4R+xd`~<lCs=9><V>E-m^rO`T1N^8U~erXTB^062t`GLTsc=ni?Di0VME^vY}A)7GyIS<cC?LP3^($NP@-rIxps|FSZCpT}0k2$2=h7D57-_6fVJz{-qtv>*I<>kryy)$$rB&>;$;M*F2W2wxEw!8LXz!q(Q@doS|f%oHJ|T2QhSZZ9C4{u!_rOF@UEuGjMz_bC$S?U^ZU5o!n4J;M$EXKn%Ni(4G)7NRiQeTVQ^02vn58DZX3AXPhJN3~8v81MJf_vF*cv)?$+ix1hm+N}mvQ>6|ZZOb~<BadM@+wjAZtP~w@7ZiFfiRS^>_`*>)x2w_`ZGw0$oTwpIIt2wb=mqh+KE@q|*hu~Z2%vK_^wje89cn0{l$B6gU{EA`bUo7s%i-I1=yPodO9uA05Fz!LcU#f*xOG7W9rGr^Hts3YWK&J7LJ?5)#RuifGU}dq}n$KY=robJ#GI9c|yKz(PgZG_fX_(B8g*GR?4$$r)9~h-NM!>4xe&J**h<^vFWrXg(OyHP-dAA2<7oW8e;jU|s>U$vzgKcuKBL10>KvJ(Oe=%g$gJZa=9Ggx0`@=hg*mCkJ$JjKR2iov`{f^-LH6g#21(5C$jeMjxzv>dTQOXcO6O>PckEP4_Q6$mqHt3j==(ND4b991DHi7q#kvQcnv8it{iuBW|u29mMY4oJGq3I%Nn#=y?t%r_?WUgeY2=C3CUbeS5X|a7@Ofh0WHBl!mRG39*a6M%xyt<6p0j3?{SYDnT&U)pWRogh%cAbohLD^FFHU(Oiw0NeF=BX39&UJ{dM}&+-!4cRjA)Os!_o!J|?ghvI^J#q+%nXhg$+0<Vo8UpfTTHBKAk50f*X{6Zp+Xbogxwj#dT&TdNX15g>HOJ3+TO4(D`#!={;U?sUu6<{LPcSFWi$k&@F^KVd}$FO+aT==mPec0i^8oix}K2I#7nv-=Q-<un>=one8<Nm%1Yke_Is_-Jold?R!kfE#sFJP)0n71ILzA!$r(?BPEbY#CEWIjcwV`BU02k_odJ#L{D>?5kWz?tiqH)Ln|EF`x97OUSAA*a^_4ZnJlLzDopbtxIYg?~$){hXP&HI%U-)$e7AT8VEFIz%9x*;FMOK81M^fbD_@gVQ68g6V+J2(^Q3UQs$KU2XeDPV0Q7qwHZq9&H<)dZW73>g%4u){qkb)GfA)Px%odDoWK0Mpg^a2N$<h}X1YlS0ChwR)8!(xId0?YY0*E3j$H5lxleoXC?t?5dH5MQDFhGGY7cvWtYBHM#D_^Wf1_C>y$O0p8Gk~^YLxg916<J3|qjv%O9jeu#LHOzGEIP7b1quV^h1iS6&2cs~km^y+q@|4F6u$4;nE)$E-b1BP}f&9^z+ukRZt~+-+!JwWz2)Xfm0?2Z+D9L3;HM>?0yoP3K{->)!;GFhbf^dq>uu-bVzsj5|tDL!i!Jzd~eaEj4(CI1n6r%*fNI(vlyta-Rli{UkWju<vK)HUhEDBQR+S{CK8|ONIwND^^VGMptruRA{BX9J?L=n>_U%<w2`Xj|)<g&deGk!uZ+AK+<Q&e$OnqmTv4rlhoL|6TnEXgof2f!FBS#HPa2#UZWg<#s|WCBOd_#0T@;>-9HFlRk4lps43NJ$BGIg!J-5Z&yTw|nxBs~5%*y9%2I9(5R=SQzvYe!rX8bcoVswm6_C2b#)9Khd+a2!e7cj<W^T1w@*|bk_!2N|FMTd+23%^M;*}9lzcm6jTA)P>9vlF6-(DiBE|lcSwjC1Bk<;A`hmDE21<L?tcj6-QQUr8VFjRJhlBR_QU_JA^+Y2>P`OZsxjB)PW0VSTEA_Rk;lA9Oq0MGKuJlVih$v^+$n{Bq3K&cwyj_$s^`98vBxOGh^c++IL!Jyv>LdPeol7dSr>VRfi8=BHaOH(ORFnIS$CET9y<r2z^DIn0;W+QRn08oNd59?Ld5nh48Q?dq_Gc}lO>kBJ7BTZU-%`);aLgp7bj5vQa3KtgJpy;QAY1CUMI&)H|~g9T)5_$^<q8r``8DZCTd`Vr+{y@H6}WBed*zqK>#FpYr@;BzmQd6<#I=JO#;~bZ9m`0x)37-0{Xieh6!5|CXPlOD8X#^cN!9t)`DUDn*2Y~vM)H0^<1_WpyDE<XE#AX_bEBr`p`5`Wn}X!JGUH>Is;APzv|1(U%5P;f}AW7Sd7nJ>*L2Iv=d+4zi!5xfaM5NlHtt2EoX<Ms)G9pli&spRhgf~;Gwe#L%AgpV`Gg4yO8Y7B_hKIl=gEt)PqJ(Y^NaNj!D$jh$Xbrr;&L(BG7y1ZC@FF;DM2P<C_ErUtgBh^h3VVdl#!A`J}|5_bId$@OWR276bf+gk0OjpsLsChw@nvb5$|N=H~BYIJ|%TAsAYMHNG<)R3t*CoxNm1C<^E&#O=2<3ge04wgauaG1*Sy4V{Wff#*3uxJMe3uBA>acHPwnjo}Xe?<8>K0!5a{i%$cb6-B@Lp32KvAO6MdCK_NnDxVaX1V<btD3c`#j$ATNyQCaAlrSktkZ!W4Rjp3EqU_z}`C5m0FH|==*R(8{0V|U@hh2<NU?!f<3wTn)E%cx?mF!Rk(#9E7d)o?@pn~=>bCQMhUlmicQ(L2$EYmt5`sI_Ya~i;`%KEBS;^&vXrvm3@h@GQvZReMU252kW9wt(gFg1^|EsJ5$DhK&u`YFwod~QHRcJHORdLdu4u+aaR#HnvIpt~zLRB9=;o%cmK5g8OsB$kPBNj@32K||d*6)O_oT4Z>9XWgqt1F_#kOE;c`(cor&zMNOxUzMf%QX|ac3u<K6(AD@ICW~t?iADO~K1}h?hsIM1P^6G{S>+P#SV7sW23Zm~1b^Lpw*L`;B46|x*dNf_hFI$_C0C@uxhslbVbyny9rR?wgZb@hphzuZ%jD&>)6bCowRc(2BOo6uK3H(tzMXaF@)_Xs4+Wt0V|NXEyQt^j?&h<p*6k`#i#LU?bwvk{xRDx3p9<nAlLRGyKZ`b{r9-HESrf5=^G_6M3@AkpQL;!34Tn@YNmes?qP;N^-6qg%=KO1()&Ma|$b~-vlwUOoHZw0>(d!6M^=*c?@<*{AASr}E6hed5USGmX(uKH$%ES*(2hQFbP)L*W`kIzZa*qe~v*$Ez_*Qb|m~J4|td(nfuJd%Z+S6yCJk23J4_H@4rpg=WAFQ&stFZF-^+Bjx^*is)7qJQZu?%<CwVj#t09n516f5wY8_g7!CA6sT;^D&c>7q_K^7`A68fd?5-`=anj8b^tHA@STq4Xwabz_2>`#<O@;izD%;7YL?CMZKm0dNvo-gsN&QQ&h6HYm>(@qgI@iK~g^ikJ3V(6|{{ZuY4p_OeSyAs4Cubl_HzK!HwO?{KET!Ro)8<RPNr3JaCcy7<*gGrw8dF?!Za46*qGEwGRKA7Am~bSbg|>uXq3OHNzBe;+WkP;fhM@rGLlY-46cNrO-pqRNP!+*&+M2MRV+C7;vCR|rc$%4&_@IpJ3*z^H=C_Ey2AAIHv##4r*RG+$K7mT2DxNbPkSwzG_3cW?gT^l>8`!Ivl<1NbiM@@gFYpVg8=l6kqimyF4m`pV7v;=VmBPO35dzq4;sK-hej4SX1&mx9e+lePhjoUZ2ZWjaqf{6uWxg2;%07L$ku)Il7<+$^0puw5NIyzIJTYHTR~0v)3Y9Zmntw5<}q+8a%siTx&E<_Akiv_4tfjj;=-V;?j7{Te{fMpt+S`VrbMjk&zbUE~3sDDZ%zrf<VYMr?0YmzfD?jl9UjfFVjQF(-V}XhbqP%wxrs<B-*~S`6@JNEt-SSm<mfHOtuN<#Tt=Qg}{<epNd6eMfQ(gF*V!Y3^CV<Rw+OO90OVI-k+iMbL`Va9-oSNwPeMBDxZo`ykSR_%_{uE(KoUdo^}*7&Wr%`Nev=a?76m)BZ(?sYb*P1)ss(6|C{$%Wz|3T`5|SBM`+A&d9mLf-MJ7vlqBk9j*lLX^%$cGmi8B!!V2JtCYkT$dhm~L7FyrUG`(ISNUf}|M}|Pn6C(oU&nhPSzE9iHXs3o>O8hn#k<~7n*%8h6Uy)@|JlPcP3QcEFm|FTrj<ksJz>Z-SoUiVz;b=aY@EdCx-0;nFoJl!(A<fjjvy(QCeG?buOZlgeL9Bbp=z`>dP8p5)5B#`5(Ua17k300UQh$Fj9nzN*Xvyu^_P8RIRQv@@YI<RIFQ!8s=nytTz_4y1EFz7$$7NWrx;ruhuCM@2+wUVspQ{iefJr=Rw!|PX1?L$S8Yat!2sff)D`XA&RWpRnFgHufu2C{ny2W{YR(Pat(a8>fPp~Ta3UV8<BB&YmkC|@Rna)6hdCGVrfy%xT*H%pOSN_L#v~P<S6yV5rm??wGQBd7eDmrOw>W2KIZeAhuc#iXAr~l-k6)S>VgNRcgna*Y4wK;f|ISb;%{%`T90C`+juNkqxZ;Un9lT%uEp|C==H(GR+<c5iyJMH=7bK5x;!m{Ssmc=83_ptIEh3&J*&>j#FS!qfS^>u$JOSH@UQ*gmOtF8BN>*n5$Ip&2ijts&`5S$F;hi`rTYA!55cHdPXDQycB#1%+rq~Tim|FVUhCPPUYbqcZD(T2g6h-LdpFbOrhp+GoU?q>p#Wr`!2f%anvOQ+gk*YeS6gjOj@_nO0N&rXdRy}5ov8<w2GBYNl=)Exdb$j2~V{ZidD;p*nipNVycbf{PNp)>kPk{Wu%Aue<F>8S@-yraHJ-o+}L-CJ{AwdfHax+KoT1S1;2oad3#uvs2fe>)R{&s%?5cHV5GCFpHvJqwu>@2CW8uD=Uwyb82qwc;xmt!d|!u}M#XfBt?(Z3s?U5By--(Ovtb=tR|m9HqM1zyKJR<8p~E|DPJw$v(GJ?PI1iqOSJyeH+0rLJ!dcs)V{gZBT<S9Bb6(Itmq-`0fqjN~VLDo{&{>#0M$a@%nL12rg3N9pWte(i9y7U)JucsjB!d;MpF9-8ccp65cR4cCeJ=`lJ!_&d9G<ZcQxK_WCmTP8`rI^J?8I-bDSKouYa#@6#`36~0oVP+Yf$D5V9Nu7O!C}CawMw(q36~Si;R;YQe*VKLb8~^|S3EXtmpgz6_00GWp?0^FRM-Y@?vBYQl0ssI200dcD')).decode()
+linecache.cache['train_gpt.py'] = (len(_src), None, _src.splitlines(True), 'train_gpt.py')
+exec(compile(_src, 'train_gpt.py', 'exec'))
diff --git a/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/train_seed1337.log b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/train_seed1337.log
new file mode 100644
index 0000000000..98c7671324
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/train_seed1337.log
@@ -0,0 +1,90 @@
+W0412 21:30:56.460000 842497 torch/distributed/run.py:851] 
+W0412 21:30:56.460000 842497 torch/distributed/run.py:851] *****************************************
+W0412 21:30:56.460000 842497 torch/distributed/run.py:851] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
+W0412 21:30:56.460000 842497 torch/distributed/run.py:851] *****************************************
+=== Direction-5: GDN Hybrid Training ===
+Arch: D_GDN_Hybrid (ARCH_MODE=D)
+Seed: 1337, Max steps: 9999, Warmdown: 1000
+Train seq_len: 2048, Wallclock budget: 590.0s
+QK_GAIN_INIT: 5.0
+World size: 8, Grad accum: 1
+EMA decay: 0.997, SWA: True (every 50)
+Late QAT threshold: 0.15
+MuonEq-R: enabled
+Validation tokens: 62,021,632
+Model built in 0.2s
+Parameters: {'embedding': 925697, 'recurrent': 13251360, 'attention': 792584, 'mlp': 18880512, 'other': 12800, 'total': 33862953}
+Total params: 33,862,953
+Matrix params: 33,263,616
+Scalar params: 75,049
+Embed params: 524,288
+Generated coprime shard order: stride across 80 shards
+
+================================================================================
+Starting training: max 9999 steps (from step 0)
+Wallclock budget: 590.0s
+================================================================================
+
+step     1/9999 | loss 6.9316 | lr_mul 0.0500 | mom 0.850 | 0.38 steps/s | 3s
+step     2/9999 | loss 6.7037 | lr_mul 0.1000 | mom 0.850 | 0.68 steps/s | 3s
+step     3/9999 | loss 6.1494 | lr_mul 0.1500 | mom 0.851 | 0.94 steps/s | 3s
+step     4/9999 | loss 5.8672 | lr_mul 0.2000 | mom 0.851 | 1.16 steps/s | 3s
+step     5/9999 | loss 5.8456 | lr_mul 0.2500 | mom 0.851 | 1.35 steps/s | 4s
+step     6/9999 | loss 5.7912 | lr_mul 0.3000 | mom 0.851 | 1.51 steps/s | 4s
+step     7/9999 | loss 5.7134 | lr_mul 0.3500 | mom 0.851 | 1.66 steps/s | 4s
+step     8/9999 | loss 5.7339 | lr_mul 0.4000 | mom 0.852 | 1.78 steps/s | 4s
+step     9/9999 | loss 5.6268 | lr_mul 0.4500 | mom 0.852 | 1.90 steps/s | 5s
+step    10/9999 | loss 5.5281 | lr_mul 0.5000 | mom 0.852 | 2.00 steps/s | 5s
+step   100/9999 | loss 3.6791 | lr_mul 1.0000 | mom 0.870 | 3.51 steps/s | 29s
+step   200/9999 | loss 2.7052 | lr_mul 1.0000 | mom 0.890 | 3.65 steps/s | 55s
+step   300/9999 | loss 2.4796 | lr_mul 1.0000 | mom 0.910 | 3.69 steps/s | 81s
+step   400/9999 | loss 2.3839 | lr_mul 1.0000 | mom 0.930 | 3.72 steps/s | 108s
+step   500/9999 | loss 2.3145 | lr_mul 1.0000 | mom 0.950 | 3.74 steps/s | 134s
+step   600/9999 | loss 2.2768 | lr_mul 1.0000 | mom 0.950 | 3.75 steps/s | 160s
+step   700/9999 | loss 2.2446 | lr_mul 1.0000 | mom 0.950 | 3.76 steps/s | 186s
+step   800/9999 | loss 2.2263 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 212s
+step   900/9999 | loss 2.2322 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 239s
+step  1000/9999 | loss 2.1975 | lr_mul 1.0000 | mom 0.950 | 3.78 steps/s | 265s
+step  1100/9999 | loss 2.1804 | lr_mul 1.0000 | mom 0.950 | 3.78 steps/s | 291s
+step  1200/9999 | loss 2.1805 | lr_mul 1.0000 | mom 0.950 | 3.78 steps/s | 317s
+step  1300/9999 | loss 2.1583 | lr_mul 1.0000 | mom 0.950 | 3.79 steps/s | 343s
+step  1400/9999 | loss 2.1301 | lr_mul 1.0000 | mom 0.950 | 3.79 steps/s | 370s
+step  1500/9999 | loss 2.1270 | lr_mul 1.0000 | mom 0.950 | 3.79 steps/s | 396s
+step  1600/9999 | loss 2.1350 | lr_mul 1.0000 | mom 0.950 | 3.79 steps/s | 422s
+step  1700/9999 | loss 2.1260 | lr_mul 1.0000 | mom 0.950 | 3.79 steps/s | 448s
+step  1800/9999 | loss 2.1136 | lr_mul 1.0000 | mom 0.950 | 3.79 steps/s | 474s
+step  1900/9999 | loss 2.1167 | lr_mul 1.0000 | mom 0.950 | 3.80 steps/s | 500s
+step  2000/9999 | loss 2.1099 | lr_mul 1.0000 | mom 0.950 | 3.80 steps/s | 527s
+step  2100/9999 | loss 2.0972 | lr_mul 1.0000 | mom 0.950 | 3.80 steps/s | 553s
+step  2200/9999 | loss 2.1001 | lr_mul 1.0000 | mom 0.950 | 3.80 steps/s | 579s
+Wallclock limit reached (590s), will stop after this step
+Stopping early at step 2242 (wallclock limit)
+
+Training complete in 590s (2242 steps)
+Peak memory: 35748 MiB
+Steps/sec: 3.80
+
+=== Applying EMA weights ===
+EMA BPB (no XSA): 1.007164
+Saved raw EMA model
+
+=== GPTQ: generating autoregressive calibration data ===
+GPTQ: generated 64 sequences, collecting hessians...
+GPTQ: collected hessians for 29 layers
+
+=== Quantizing to int6 + zstd-22 ===
+Artifact: 15,903,365 bytes (15.17 MB)
+
+=== Roundtrip Validation (quantized model) ===
+Quantized BPB (no XSA): 1.015700
+Quantization degradation: +0.008536
+Quantized BPB (XSA-all): 1.020105
+
+================================================================================
+FINAL RESULTS — D_GDN_Hybrid seed=1337
+  Training: 2242 steps, 590s
+  EMA BPB:       1.007164
+  Quantized BPB: 1.015700
+  XSA BPB:       1.020105
+  Artifact:      /root/pg-repo/records/track_10min_16mb/2026-04-12_JM_GDN_Hybrid_Warmdown1000_CompressedCode_Seeds42_1337_2024/checkpoints/final_model_D_GDN_Hybrid_seed1337.int6.ptz
+================================================================================
diff --git a/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/train_seed2024.log b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/train_seed2024.log
new file mode 100644
index 0000000000..fe4f609d72
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/train_seed2024.log
@@ -0,0 +1,90 @@
+W0412 21:55:57.316000 848542 torch/distributed/run.py:851] 
+W0412 21:55:57.316000 848542 torch/distributed/run.py:851] *****************************************
+W0412 21:55:57.316000 848542 torch/distributed/run.py:851] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
+W0412 21:55:57.316000 848542 torch/distributed/run.py:851] *****************************************
+=== Direction-5: GDN Hybrid Training ===
+Arch: D_GDN_Hybrid (ARCH_MODE=D)
+Seed: 2024, Max steps: 9999, Warmdown: 1000
+Train seq_len: 2048, Wallclock budget: 590.0s
+QK_GAIN_INIT: 5.0
+World size: 8, Grad accum: 1
+EMA decay: 0.997, SWA: True (every 50)
+Late QAT threshold: 0.15
+MuonEq-R: enabled
+Validation tokens: 62,021,632
+Model built in 0.2s
+Parameters: {'embedding': 925697, 'recurrent': 13251360, 'attention': 792584, 'mlp': 18880512, 'other': 12800, 'total': 33862953}
+Total params: 33,862,953
+Matrix params: 33,263,616
+Scalar params: 75,049
+Embed params: 524,288
+Generated coprime shard order: stride across 80 shards
+
+================================================================================
+Starting training: max 9999 steps (from step 0)
+Wallclock budget: 590.0s
+================================================================================
+
+step     1/9999 | loss 6.9320 | lr_mul 0.0500 | mom 0.850 | 0.37 steps/s | 3s
+step     2/9999 | loss 6.7120 | lr_mul 0.1000 | mom 0.850 | 0.67 steps/s | 3s
+step     3/9999 | loss 6.1764 | lr_mul 0.1500 | mom 0.851 | 0.92 steps/s | 3s
+step     4/9999 | loss 5.8613 | lr_mul 0.2000 | mom 0.851 | 1.14 steps/s | 4s
+step     5/9999 | loss 5.8159 | lr_mul 0.2500 | mom 0.851 | 1.33 steps/s | 4s
+step     6/9999 | loss 5.7956 | lr_mul 0.3000 | mom 0.851 | 1.49 steps/s | 4s
+step     7/9999 | loss 5.7609 | lr_mul 0.3500 | mom 0.851 | 1.63 steps/s | 4s
+step     8/9999 | loss 5.7081 | lr_mul 0.4000 | mom 0.852 | 1.76 steps/s | 5s
+step     9/9999 | loss 5.6193 | lr_mul 0.4500 | mom 0.852 | 1.87 steps/s | 5s
+step    10/9999 | loss 5.5309 | lr_mul 0.5000 | mom 0.852 | 1.97 steps/s | 5s
+step   100/9999 | loss 3.6509 | lr_mul 1.0000 | mom 0.870 | 3.50 steps/s | 29s
+step   200/9999 | loss 2.7050 | lr_mul 1.0000 | mom 0.890 | 3.64 steps/s | 55s
+step   300/9999 | loss 2.4970 | lr_mul 1.0000 | mom 0.910 | 3.68 steps/s | 81s
+step   400/9999 | loss 2.4011 | lr_mul 1.0000 | mom 0.930 | 3.71 steps/s | 108s
+step   500/9999 | loss 2.3435 | lr_mul 1.0000 | mom 0.950 | 3.73 steps/s | 134s
+step   600/9999 | loss 2.3035 | lr_mul 1.0000 | mom 0.950 | 3.75 steps/s | 160s
+step   700/9999 | loss 2.2498 | lr_mul 1.0000 | mom 0.950 | 3.75 steps/s | 186s
+step   800/9999 | loss 2.2275 | lr_mul 1.0000 | mom 0.950 | 3.76 steps/s | 213s
+step   900/9999 | loss 2.2170 | lr_mul 1.0000 | mom 0.950 | 3.76 steps/s | 239s
+step  1000/9999 | loss 2.1922 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 265s
+step  1100/9999 | loss 2.1769 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 292s
+step  1200/9999 | loss 2.1650 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 318s
+step  1300/9999 | loss 2.1550 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 345s
+step  1400/9999 | loss 2.1450 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 371s
+step  1500/9999 | loss 2.1311 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 397s
+step  1600/9999 | loss 2.1362 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 424s
+step  1700/9999 | loss 2.1506 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 450s
+step  1800/9999 | loss 2.1119 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 477s
+step  1900/9999 | loss 2.1248 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 503s
+step  2000/9999 | loss 2.1295 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 530s
+step  2100/9999 | loss 2.1179 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 557s
+step  2200/9999 | loss 2.1044 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 583s
+Wallclock limit reached (590s), will stop after this step
+Stopping early at step 2227 (wallclock limit)
+
+Training complete in 590s (2227 steps)
+Peak memory: 35748 MiB
+Steps/sec: 3.77
+
+=== Applying EMA weights ===
+EMA BPB (no XSA): 1.009032
+Saved raw EMA model
+
+=== GPTQ: generating autoregressive calibration data ===
+GPTQ: generated 64 sequences, collecting hessians...
+GPTQ: collected hessians for 29 layers
+
+=== Quantizing to int6 + zstd-22 ===
+Artifact: 15,713,422 bytes (14.99 MB)
+
+=== Roundtrip Validation (quantized model) ===
+Quantized BPB (no XSA): 1.018237
+Quantization degradation: +0.009205
+Quantized BPB (XSA-all): 1.024111
+
+================================================================================
+FINAL RESULTS — D_GDN_Hybrid seed=2024
+  Training: 2227 steps, 590s
+  EMA BPB:       1.009032
+  Quantized BPB: 1.018237
+  XSA BPB:       1.024111
+  Artifact:      /root/pg-repo/records/track_10min_16mb/2026-04-12_JM_GDN_Hybrid_Warmdown1000_CompressedCode_Seeds42_1337_2024/checkpoints/final_model_D_GDN_Hybrid_seed2024.int6.ptz
+================================================================================
diff --git a/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/train_seed42.log b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/train_seed42.log
new file mode 100644
index 0000000000..234b80f64a
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-12_GDN_Hybrid_DeltaRule_Warmdown1000_CompressedCode_1.0167/train_seed42.log
@@ -0,0 +1,90 @@
+W0412 21:05:55.874000 839649 torch/distributed/run.py:851] 
+W0412 21:05:55.874000 839649 torch/distributed/run.py:851] *****************************************
+W0412 21:05:55.874000 839649 torch/distributed/run.py:851] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
+W0412 21:05:55.874000 839649 torch/distributed/run.py:851] *****************************************
+=== Direction-5: GDN Hybrid Training ===
+Arch: D_GDN_Hybrid (ARCH_MODE=D)
+Seed: 42, Max steps: 9999, Warmdown: 1000
+Train seq_len: 2048, Wallclock budget: 590.0s
+QK_GAIN_INIT: 5.0
+World size: 8, Grad accum: 1
+EMA decay: 0.997, SWA: True (every 50)
+Late QAT threshold: 0.15
+MuonEq-R: enabled
+Validation tokens: 62,021,632
+Model built in 0.2s
+Parameters: {'embedding': 925697, 'recurrent': 13251360, 'attention': 792584, 'mlp': 18880512, 'other': 12800, 'total': 33862953}
+Total params: 33,862,953
+Matrix params: 33,263,616
+Scalar params: 75,049
+Embed params: 524,288
+Generated coprime shard order: stride across 80 shards
+
+================================================================================
+Starting training: max 9999 steps (from step 0)
+Wallclock budget: 590.0s
+================================================================================
+
+step     1/9999 | loss 6.9368 | lr_mul 0.0500 | mom 0.850 | 0.38 steps/s | 3s
+step     2/9999 | loss 6.7108 | lr_mul 0.1000 | mom 0.850 | 0.68 steps/s | 3s
+step     3/9999 | loss 6.1586 | lr_mul 0.1500 | mom 0.851 | 0.94 steps/s | 3s
+step     4/9999 | loss 5.8825 | lr_mul 0.2000 | mom 0.851 | 1.16 steps/s | 3s
+step     5/9999 | loss 5.8143 | lr_mul 0.2500 | mom 0.851 | 1.34 steps/s | 4s
+step     6/9999 | loss 5.7683 | lr_mul 0.3000 | mom 0.851 | 1.51 steps/s | 4s
+step     7/9999 | loss 5.7302 | lr_mul 0.3500 | mom 0.851 | 1.65 steps/s | 4s
+step     8/9999 | loss 5.7048 | lr_mul 0.4000 | mom 0.852 | 1.77 steps/s | 5s
+step     9/9999 | loss 5.6385 | lr_mul 0.4500 | mom 0.852 | 1.89 steps/s | 5s
+step    10/9999 | loss 5.5578 | lr_mul 0.5000 | mom 0.852 | 1.99 steps/s | 5s
+step   100/9999 | loss 3.6592 | lr_mul 1.0000 | mom 0.870 | 3.48 steps/s | 29s
+step   200/9999 | loss 2.6998 | lr_mul 1.0000 | mom 0.890 | 3.63 steps/s | 55s
+step   300/9999 | loss 2.5020 | lr_mul 1.0000 | mom 0.910 | 3.67 steps/s | 82s
+step   400/9999 | loss 2.4033 | lr_mul 1.0000 | mom 0.930 | 3.70 steps/s | 108s
+step   500/9999 | loss 2.3428 | lr_mul 1.0000 | mom 0.950 | 3.72 steps/s | 134s
+step   600/9999 | loss 2.3076 | lr_mul 1.0000 | mom 0.950 | 3.73 steps/s | 161s
+step   700/9999 | loss 2.2690 | lr_mul 1.0000 | mom 0.950 | 3.74 steps/s | 187s
+step   800/9999 | loss 2.2464 | lr_mul 1.0000 | mom 0.950 | 3.74 steps/s | 214s
+step   900/9999 | loss 2.2002 | lr_mul 1.0000 | mom 0.950 | 3.75 steps/s | 240s
+step  1000/9999 | loss 2.1897 | lr_mul 1.0000 | mom 0.950 | 3.75 steps/s | 266s
+step  1100/9999 | loss 2.1657 | lr_mul 1.0000 | mom 0.950 | 3.76 steps/s | 293s
+step  1200/9999 | loss 2.1626 | lr_mul 1.0000 | mom 0.950 | 3.76 steps/s | 319s
+step  1300/9999 | loss 2.1452 | lr_mul 1.0000 | mom 0.950 | 3.76 steps/s | 346s
+step  1400/9999 | loss 2.1373 | lr_mul 1.0000 | mom 0.950 | 3.76 steps/s | 372s
+step  1500/9999 | loss 2.1262 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 398s
+step  1600/9999 | loss 2.1349 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 425s
+step  1700/9999 | loss 2.1228 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 451s
+step  1800/9999 | loss 2.1123 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 477s
+step  1900/9999 | loss 2.1054 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 504s
+step  2000/9999 | loss 2.1007 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 530s
+step  2100/9999 | loss 2.1106 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 557s
+step  2200/9999 | loss 2.1059 | lr_mul 1.0000 | mom 0.950 | 3.77 steps/s | 583s
+Wallclock limit reached (590s), will stop after this step
+Stopping early at step 2227 (wallclock limit)
+
+Training complete in 590s (2227 steps)
+Peak memory: 35748 MiB
+Steps/sec: 3.77
+
+=== Applying EMA weights ===
+EMA BPB (no XSA): 1.007164
+Saved raw EMA model
+
+=== GPTQ: generating autoregressive calibration data ===
+GPTQ: generated 64 sequences, collecting hessians...
+GPTQ: collected hessians for 29 layers
+
+=== Quantizing to int6 + zstd-22 ===
+Artifact: 15,733,879 bytes (15.00 MB)
+
+=== Roundtrip Validation (quantized model) ===
+Quantized BPB (no XSA): 1.016200
+Quantization degradation: +0.009035
+Quantized BPB (XSA-all): 1.021202
+
+================================================================================
+FINAL RESULTS — D_GDN_Hybrid seed=42
+  Training: 2227 steps, 590s
+  EMA BPB:       1.007164
+  Quantized BPB: 1.016200
+  XSA BPB:       1.021202
+  Artifact:      /root/pg-repo/records/track_10min_16mb/2026-04-12_JM_GDN_Hybrid_Warmdown1000_CompressedCode_Seeds42_1337_2024/checkpoints/final_model_D_GDN_Hybrid_seed42.int6.ptz
+================================================================================