id stringlengths 36 36 | tier stringclasses 1
value | github_login stringclasses 179
values | created_at stringlengths 32 32 | status stringclasses 2
values | sha256 stringlengths 64 64 | source_bytes int64 341 260k | leaderboard_rank int64 | source large_stringlengths 341 260k | metadata_json large_stringlengths 645 734 | result_json large_stringlengths 5 4.8k |
|---|---|---|---|---|---|---|---|---|---|---|
17dc9a40-db63-43e0-8815-eb9d8d5682f4 | easy | yashkant | 2026-08-15 14:19:50.231542+00:00 | succeeded | e2494728ba6de1810b534646ddb968eeab5904099d8673c05951a2cf883fbbc9 | 35,314 | null | """Round1232 parsed-T allocation crossover on the y/z MinMax transition.
The complete Round1220 answer/latent MinMax candidate remains unchanged except
for the public-T-to-macro allocation. Separate answer y and latent z states,
immutable x reinjection, both MinMax updates per macro, and full BPTT remain
intact. MINMA... | {
"id": "17dc9a40-db63-43e0-8815-eb9d8d5682f4",
"created_at": "2026-08-15 14:19:50.231542+00:00",
"db_md5": "39df511e10424a63a2bf7764afd6af54",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "661657b2-4349-4db5-aacf-784ed9384d1a",
"tier": "easy",
"dataset_id": "e1",
"status": "succeed... | {
"score": {
"mean_loss": 4.702207088470459,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.08666666597127914
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 5.617396831512451,
"example_count": 100,
... |
17e09af2-0381-47b3-b56d-8f38726c35f2 | easy | AFeinfield | 2026-08-18 12:06:40.832683+00:00 | succeeded | 5a1903ef6f436145efe2e10e10cb771ba4750e70fafbe814af4e4d2bb3f16693 | 13,045 | null | """Looped factored block (055) + K=5 output-head agreement ensemble.
065: identical to 055 except the single tied output head becomes five:
head 0 keeps the token-embedding tie, plus four independent
Linear(d, vocab, bias=False) heads (normal init, std 0.02).
* TRAINING: return the MEAN of the five heads' logits. T... | {
"id": "17e09af2-0381-47b3-b56d-8f38726c35f2",
"created_at": "2026-08-18 12:06:40.832683+00:00",
"db_md5": "4bdfc63619ebb62fe644a9f159981e80",
"submitter": "AFeinfield",
"github_login": "AFeinfield",
"run_id": "833bebf3-474e-4748-b798-d50f6458b964",
"tier": "easy",
"dataset_id": "e6",
"status": "succ... | {
"score": {
"mean_loss": 12.334324836730957,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.05180180352181196
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 13.118236541748047,
"example_count": 60,... |
17e0b4c3-f6df-4c75-acd4-43bc18f51f95 | easy | AFeinfield | 2026-08-18 23:46:28.926081+00:00 | succeeded | 6bb044a895217f39cca056017c0554724f648c1c554b178429dd30e096c0f86b | 13,734 | null | """084: wide conv (kernel 9) + carry-GRU removed — the two best
fit-speed findings combined. 083 hit 100% train on e2; 081 gave +20%
steps/s. Generalization must be re-checked on m6.
Original 083 notes follow.
083: conv kernel 5 -> 9 in the multiply sublayer.
Direct test of the mechanism story: the conv gathers neighb... | {
"id": "17e0b4c3-f6df-4c75-acd4-43bc18f51f95",
"created_at": "2026-08-18 23:46:28.926081+00:00",
"db_md5": "4c146f4e4f1b8669095b715d12011ce4",
"submitter": "AFeinfield",
"github_login": "AFeinfield",
"run_id": "4276d7f2-24a1-47f6-a953-15841b7f8dc7",
"tier": "easy",
"dataset_id": "e2",
"status": "succ... | {
"score": {
"mean_loss": 16.343594551086426,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.01416666666045785
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 17.00062370300293,
"example_count": 300,... |
17e1bbc4-5ab0-41b3-b57f-ad607321c28f | easy | Yalyenea | 2026-08-12 06:13:36.898487+00:00 | succeeded | 21142510da6732cff94b031cb37f894b228da151d964aea4c276aea60f53fa95 | 10,560 | null | """Adaptive recurrent decimal-slot state with a direct causal suffix readout."""
from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
assert_model_st... | {
"id": "17e1bbc4-5ab0-41b3-b57f-ad607321c28f",
"created_at": "2026-08-12 06:13:36.898487+00:00",
"db_md5": "893882b608f242ebe53383009b681233",
"submitter": "yfff",
"github_login": "Yalyenea",
"run_id": "a8e633c3-38d7-4cd2-8668-f70e3dde8b51",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded",
... | {
"score": {
"mean_loss": 2.2171345189347997,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.008333333395421505
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.218374490737915,
"example_count": 600... |
17e33ba8-6ffe-4753-948b-115c69d0daed | easy | karanganesan | 2026-08-04 02:03:20.944591+00:00 | succeeded | a9f40f8a9044651de96df9da960df48e0e547f70ca238a374ae59371c7c514b0 | 11,979 | null | """Parametric looped-transformer family (P1).
One weight-tied transformer block applied k times in latent space. Config
flags cover four P1 families with one file:
looped recall=0 gated=0 tfilm=0 plain weight-tied loop
looped-recall recall=1 re-inject the input embedding each
... | {
"id": "17e33ba8-6ffe-4753-948b-115c69d0daed",
"created_at": "2026-08-04 02:03:20.944591+00:00",
"db_md5": "e6071d7335a82dfd09d1a16f8191eeb3",
"submitter": "Karan Ganesan",
"github_login": "karanganesan",
"run_id": "2177ced8-c2da-412c-8607-f4d5be590825",
"tier": "easy",
"dataset_id": "e1",
"status": ... | {
"score": {
"mean_loss": 2.5842727422714233,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.021666666492819786
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.205512285232544,
"example_count": 100... |
17e3979e-30e7-47c5-8f9c-57b17a675aed | easy | shirvani-jr | 2026-08-21 21:53:25.453513+00:00 | succeeded | 5e2794afc35c27dfbc3a7eefb4bba09c80dff190e2475b7f7bfec2f2e2220732 | 12,088 | null | """Looped digit-state squaring machine.
One recurrent application of a shared block corresponds to one modular
squaring. The recurrent state is the decimal digit string of the current
value, re-quantised every step with a straight-through estimator, so the
learned map composes exactly and depth extrapolates in T.
"""... | {
"id": "17e3979e-30e7-47c5-8f9c-57b17a675aed",
"created_at": "2026-08-21 21:53:25.453513+00:00",
"db_md5": "4aec1e262dd141b8cb993fb7a46c23f9",
"submitter": "Ali",
"github_login": "shirvani-jr",
"run_id": "c3396b9a-d034-4ba2-9f2d-7f07387af898",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.468626379966736,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.5883333384990692
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.5449833869934082,
"example_count": 100,
... |
17e3c6ac-c906-4231-83cd-624de9597e78 | easy | DDanlov | 2026-08-13 22:55:08.400822+00:00 | failed | e560401ef61a30a9c3eee7e603d1d2a77b67632ca4829f4d485fd63743e6db79 | 22,441 | null | from __future__ import annotations
import torch
import torch.nn as nn
import torch.nn.functional as F
import math
from dataclasses import dataclass
from typing import Optional
from torch.optim import Optimizer
try:
from client import Submission, assert_model_state, OptimizerBundle
except ImportError:
Submissio... | {
"id": "17e3c6ac-c906-4231-83cd-624de9597e78",
"created_at": "2026-08-13 22:55:08.400822+00:00",
"db_md5": "21142b4474ca7f0074240c8006f5098e",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "30072df3-827c-4916-8da5-01b22df6548f",
"tier": "easy",
"dataset_id": "e5",
"status": "failed",
... | null
|
17e40a97-ba62-4b5f-8789-46e7c64222dd | easy | oupadhyay | 2026-08-07 10:29:41.719670+00:00 | succeeded | a91d745c50a5f327f24c00aa7c30a8ed2e2755cbb93d387480ff91275b0396b6 | 5,484 | null | """Baseline AdamW Transformer with the CUDA-safe compressed-axial v3 parser."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
assert_model_state,
)
PAD, ... | {
"id": "17e40a97-ba62-4b5f-8789-46e7c64222dd",
"created_at": "2026-08-07 10:29:41.719670+00:00",
"db_md5": "60704a6bd82bd4323511bb446819c524",
"submitter": "Ojasw Upadhyay",
"github_login": "oupadhyay",
"run_id": "4785888f-4759-4cfc-a02a-c4c425a4382e",
"tier": "easy",
"dataset_id": "e5",
"status": "s... | {
"score": {
"mean_loss": 74.31122951336025,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0004166666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 75.17941961160155,
"example_count": 600... |
17e5c4fa-3957-43b4-b475-8dfc894a197e | easy | anirudh-chakravarthy | 2026-08-12 04:07:22.280879+00:00 | succeeded | 17653d4b1a1eb44b0f60a4133ba157fb5984a5bacc2f3f7ff0c150b6625afd46 | 15,334 | null | """A learned state transition applied once per requested time step."""
from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
asser... | {
"id": "17e5c4fa-3957-43b4-b475-8dfc894a197e",
"created_at": "2026-08-12 04:07:22.280879+00:00",
"db_md5": "82a4f87339093d3157461322b9cda3ba",
"submitter": "Anirudh S Chakravarthy",
"github_login": "anirudh-chakravarthy",
"run_id": "1e4487d0-a40e-469b-adfb-df56875d16ba",
"tier": "easy",
"dataset_id": "... | {
"score": {
"mean_loss": 3.9867569979796293,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.011666666753590106
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.854446094666896,
"example_count": 600... |
17e5d319-4a58-4772-a8c2-d4ba73426127 | easy | khushidahi | 2026-08-22 01:41:56.389063+00:00 | failed | 6ce8a5e14101de39e2f7bf59528bb3ae02221e7bffabec0a02fe1c1d58dbfc47 | 19,188 | null | """R395: minimal Universal-Transformer functional-root control for M5.
This candidate adapts the released Universal Transformer to the competition's
separate-output interface. Numeric prompt symbols are only rearranged onto
equal-length, right-aligned categorical N and X tapes. A third tape supplies
parallel learned... | {
"id": "17e5d319-4a58-4772-a8c2-d4ba73426127",
"created_at": "2026-08-22 01:41:56.389063+00:00",
"db_md5": "7cc7dd5ec02477b658d8a7bd866b93ec",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "5ff85aff-dcfb-4184-bd4f-8e1b95683b01",
"tier": "easy",
"dataset_id": "e5",
"status": "fail... | null
|
17f14a31-3884-47d7-b0d3-933c558d6003 | easy | GLauzza | 2026-08-22 10:24:44.120577+00:00 | succeeded | 7bf30e19d90e7f6cadb4ad67d0b2cd51634a4fd2d29a507be6f840a0dd78f8c1 | 3,628 | null | """Basic single-pass Transformer with PyTorch AdamW."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
assert_model_state,
)
D_MODEL = 128
NUM_HEADS = 4
... | {
"id": "17f14a31-3884-47d7-b0d3-933c558d6003",
"created_at": "2026-08-22 10:24:44.120577+00:00",
"db_md5": "eaac69fe345f82973fcf392c504a40ff",
"submitter": "Gabriel",
"github_login": "GLauzza",
"run_id": "872daebf-710a-4834-bd8f-3df7a49e3ce5",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 3.0526736974716187,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.018333333544433117
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.00348424911499,
"example_count": 100,... |
1802f5ac-a0eb-4463-b794-0a0d52ae0041 | easy | yashkant | 2026-08-09 06:11:06.232821+00:00 | succeeded | 635fe72b9c33d000d7535ea411771d815ef1ac1b946671a345dd23311cf0a3c3 | 19,915 | null | """Round 305 depth-3 parent with two softly T-gated shared refinements."""
from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
assert_model_state,
)
from torc... | {
"id": "1802f5ac-a0eb-4463-b794-0a0d52ae0041",
"created_at": "2026-08-09 06:11:06.232821+00:00",
"db_md5": "3d87980149fee64d23863e4fe6340e66",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "a6f8d09f-c64d-49dc-bf06-fe3241e99c2e",
"tier": "easy",
"dataset_id": "e2",
"status": "succeed... | {
"score": {
"mean_loss": 4.305101811885834,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.3477083444595337
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 7.1622419357299805,
"example_count": 300,
... |
180b4bfa-efa4-4151-af6e-476e292b130b | easy | blake-camp-surge | 2026-08-04 17:15:16.661077+00:00 | succeeded | 2de3ab313754416d1bbf2003a63cb71b6313fe883ef29aa66d3cb69e93b3aaa1 | 15,402 | null | """Recurrent Transformer with residual DAN-generated hyper-attention."""
from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
ass... | {
"id": "180b4bfa-efa4-4151-af6e-476e292b130b",
"created_at": "2026-08-04 17:15:16.661077+00:00",
"db_md5": "cee1cfd65e9a1d18352c99b8fa79d1cb",
"submitter": "Blake Camp",
"github_login": "blake-camp-surge",
"run_id": "6c4ffe2f-4b39-42d1-9a4c-3d3c1021adc1",
"tier": "easy",
"dataset_id": "e3",
"status":... | {
"score": {
"mean_loss": 5.477224101513071,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.006875
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 5.440849859673384,
"example_count": 800,
"... |
180ebdd1-2ec3-4620-968e-f7b02a7944f4 | easy | yashkant | 2026-08-12 00:12:40.911342+00:00 | succeeded | 6e0d0dbd43d425640778754e09254021d887e51849f3b11f65153d1acc0ecd00 | 31,388 | null | """Round1702 one-projection-per-T Dykstra recursion on the D416 carrier.
Each parsed public-T cycle applies one shared learned smooth projection and
alternates the A/B set code and correction memory across cycles. The immutable
anchor is reinjected on every call; all active cycle outputs are averaged, and
terminal CE... | {
"id": "180ebdd1-2ec3-4620-968e-f7b02a7944f4",
"created_at": "2026-08-12 00:12:40.911342+00:00",
"db_md5": "de8307260393c74454c1cc99e1c8fe9b",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "ec54258e-46ef-4428-9a98-722a81b55070",
"tier": "easy",
"dataset_id": "e3",
"status": "succeed... | {
"score": {
"mean_loss": 2.123751995135529,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.01875000004656613
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.121284371321072,
"example_count": 800,
... |
18128378-ddf5-4d7f-990a-6024a57775cc | easy | benjaminW2025 | 2026-08-20 22:49:50.789951+00:00 | succeeded | 5eb6caedc32c58c030d5a37fee22da5b85bd41ba5f0f7b6c5980bbee4fa17545 | 5,487 | null | """Phase 1 U8: eight untied blocks with hard-sequence weighted CE."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
assert_model_state,... | {
"id": "18128378-ddf5-4d7f-990a-6024a57775cc",
"created_at": "2026-08-20 22:49:50.789951+00:00",
"db_md5": "ad3420bbe7e5447343c7bf1a95aff583",
"submitter": "benjawesome",
"github_login": "benjaminW2025",
"run_id": "ec859d2e-7ec3-48aa-836e-a02c11ea1d95",
"tier": "easy",
"dataset_id": "e3",
"status": "... | {
"score": {
"mean_loss": 2.189536138182519,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.011250000013969839
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.235916720552007,
"example_count": 800,... |
181401f7-5532-4118-ac6c-9618d2986971 | easy | velocizapkar | 2026-08-30 16:42:21.157083+00:00 | succeeded | 5c8ee47bbc801493bd7ba5eb52143381406519f108fd1d2986fa986067fa1ac7 | 7,486 | null | """HeightGrid: a (sequence x depth x height) transformer for One Layer Deeper.
The computation is a 3D grid. Each cell (t, d, h) holds a width-W vector and
receives three interactions:
* sequence: softmax attention over cells (., d, h) in the same plane;
* height: the running state ascending from (t, d, h-1);
... | {
"id": "181401f7-5532-4118-ac6c-9618d2986971",
"created_at": "2026-08-30 16:42:21.157083+00:00",
"db_md5": "60f12ad7eb595dacf72033c30dfd11dd",
"submitter": "Aakanksh Zarapkar",
"github_login": "velocizapkar",
"run_id": "f82e163c-e466-497c-859c-97147f070af5",
"tier": "easy",
"dataset_id": "e4",
"statu... | {
"score": {
"mean_loss": 2.1075797401625826,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.003333333333333333
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1456088605186032,
"example_count": 12... |
1815cdc8-39a0-4262-9655-e4d75487d64c | easy | AFeinfield | 2026-08-28 04:04:47.895628+00:00 | succeeded | 75e1fd8d7df9293d7a9a7a7d7ffbf59a4d7a1d2e48d93476b0a008ee97ce6b39 | 11,546 | null | """184: 182 recipe arm gf_lamb=1.0 wd=1.0.
182 notes: SPELL-HEAD AUX (head learns to spell every residue seen
as an input, from the prompt's own digit tokens — pure re-encoding).
Targets the residual ~15%: held-out answers whose residues never appear
as training answers were unspellable however well F routed.
Local p... | {
"id": "1815cdc8-39a0-4262-9655-e4d75487d64c",
"created_at": "2026-08-28 04:04:47.895628+00:00",
"db_md5": "30c738e1e5b776c8e74adc8c55e864b7",
"submitter": "AFeinfield",
"github_login": "AFeinfield",
"run_id": "c5bb3c83-500e-4b84-94fc-fa23f707b89f",
"tier": "easy",
"dataset_id": "e3",
"status": "succ... | {
"score": {
"mean_loss": 2.640986086811718,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.026250000060535968
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.6158634424711105,
"example_count": 800... |
1817a565-c3dc-4d2c-8a64-3cee662824b7 | easy | DDanlov | 2026-08-12 20:17:43.892495+00:00 | failed | 5e2371dff926bd6839dee3b4b9860f9b884a2dfd5b6030fe6572b7b9278ceb45 | 25,988 | null | from __future__ import annotations
import math
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim.optimizer import Optimizer
try:
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
assert_model_state,
)
except ... | {
"id": "1817a565-c3dc-4d2c-8a64-3cee662824b7",
"created_at": "2026-08-12 20:17:43.892495+00:00",
"db_md5": "9071fd4645ee4c014d5cbdb3e9321ead",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "5ac8bea9-4200-449d-bf89-a7e1c54ff997",
"tier": "easy",
"dataset_id": "e4",
"status": "failed",
... | null
|
181d6831-d9e3-45d9-8c87-34f0ef00cfba | easy | SunstarKK | 2026-08-18 13:26:10.302678+00:00 | succeeded | 0a94c24689544ee32b9529047c8e7a0f76246491d3a228f70358ddf4af98205c | 20,414 | null | """Heterogeneous transition with natural successor supervision."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
assert_model_state,
)
... | {
"id": "181d6831-d9e3-45d9-8c87-34f0ef00cfba",
"created_at": "2026-08-18 13:26:10.302678+00:00",
"db_md5": "61cc515aa8486019980cf98193ef6733",
"submitter": "SunstarK",
"github_login": "SunstarKK",
"run_id": "4287f9ee-d5c7-4b1a-91e9-98abb074e078",
"tier": "easy",
"dataset_id": "e5",
"status": "succeed... | {
"score": {
"mean_loss": 2.245857952045159,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.011249999900658926
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.175225019454956,
"example_count": 600,... |
1822cf89-5c2d-410e-a804-4be2849de59f | easy | briheet | 2026-08-07 01:36:40.633861+00:00 | succeeded | dc731dea9b474a0a6bf011028c00e4f9983e0a62e8c1d4fba26fab6380db237b | 3,932 | null | """Basic single-pass Transformer with PyTorch AdamW."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
assert_model_state,
)
D_MODEL = 96
NUM_HEADS = 4
c... | {
"id": "1822cf89-5c2d-410e-a804-4be2849de59f",
"created_at": "2026-08-07 01:36:40.633861+00:00",
"db_md5": "e42602028d70eb81ebf749b63cb042ac",
"submitter": "Aizenn",
"github_login": "briheet",
"run_id": "1a2d54c8-df2e-4ff7-8526-f8171b7a09b9",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded",... | {
"score": {
"mean_loss": 5.085044622421265,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.04166666604578495
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 5.351727485656738,
"example_count": 100,
... |
1828c269-b3a7-478c-b759-9fcfd564ff6a | easy | erdavis0 | 2026-08-26 04:28:41.922927+00:00 | succeeded | b063dd87e5c8ed86565b7dc9bb9038c969fb662780579ead5b62238da52e334a | 30,346 | null | """Latent-role continuous REV control C0: ordinary evaluator token CE.
Public delimiter syntax only routes opaque token categories into distinct,
right-aligned N, X, and T tapes. Learned local cells compile separate
interaction, carry, and modulus-conditioned workspaces. N is isolated from
interaction and carry and ... | {
"id": "1828c269-b3a7-478c-b759-9fcfd564ff6a",
"created_at": "2026-08-26 04:28:41.922927+00:00",
"db_md5": "b092954054ec5242242d7829dc8330c1",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "925e940c-f987-476d-96c9-bf0301c70ef6",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded",... | {
"score": {
"mean_loss": 4.238319516181946,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.06833333522081375
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.523534774780273,
"example_count": 100,
... |
1832ed86-8518-45aa-9202-ee29352f8122 | easy | yashkant | 2026-08-13 01:00:00.122895+00:00 | succeeded | ff811309ab88122598995c3a05873c8fcc7ace221390797563883508229db4b6 | 31,750 | null | """Round1701 Dykstra width16/hidden32 architecture pair on the D416 MemGLU carrier.
The fixed Round1141 carrier first produces its ordinary pooled representation.
A compact primal state and two Dykstra correction memories are initialized
from that immutable representation. Every parsed public-T cycle applies two
cond... | {
"id": "1832ed86-8518-45aa-9202-ee29352f8122",
"created_at": "2026-08-13 01:00:00.122895+00:00",
"db_md5": "fe6945073e52cb66bf9db39d3c089c6d",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "57a16368-a26b-4ec5-aebb-51f3079df30a",
"tier": "easy",
"dataset_id": "e4",
"status": "succeed... | {
"score": {
"mean_loss": 2.0647870608167556,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.010729166679084301
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.0672343179962205,
"example_count": 12... |
1838ca1b-1234-47b0-bb57-56af8dad0daf | easy | DDanlov | 2026-08-28 22:11:49.250372+00:00 | succeeded | e39fe62a52d2461b6f44d78c5571b356101728c44efdb52869b4c54f4fad2008 | 15,430 | null | """
Trial 36: Model D: D=256, H=4, M=8, u=2 (T=16)
"""
import math
import torch
import torch.nn as nn
import torch.nn.functional as F
from dataclasses import dataclass
from typing import Optional, Tuple, Dict, Any, List, Union
try:
from benchmark.api import ModelSpec, OptimizerBundle, OptimizerSpec, Submission, T... | {
"id": "1838ca1b-1234-47b0-bb57-56af8dad0daf",
"created_at": "2026-08-28 22:11:49.250372+00:00",
"db_md5": "b25e9f6d88d6211fb55585dc8d30d9c2",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "e1a83c22-a7ff-4d75-8e97-77a77492d758",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.5179332400630487,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.009166666679084301
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.5388761073217263,
"example_count": 60... |
183a836c-f073-49ea-8305-2dfb2869a48d | easy | karanganesan | 2026-08-07 07:59:16.630385+00:00 | succeeded | a31ed38f7c55e098b584fe8f7b36db50f6fefdd6e951bb04a85425578c4aa65d | 25,600 | null | """Parametric looped-transformer family (P1).
One weight-tied transformer block applied k times in latent space. Config
flags cover four P1 families with one file:
looped recall=0 gated=0 tfilm=0 plain weight-tied loop
looped-recall recall=1 re-inject the input embedding each
... | {
"id": "183a836c-f073-49ea-8305-2dfb2869a48d",
"created_at": "2026-08-07 07:59:16.630385+00:00",
"db_md5": "de878c6992e9f0d2d1e563bfe42f905e",
"submitter": "Karan Ganesan",
"github_login": "karanganesan",
"run_id": "5344bc8e-0101-4fc9-aead-d326cdfe809d",
"tier": "easy",
"dataset_id": "e5",
"status": ... | {
"score": {
"mean_loss": 3.894197940826416,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0070833335048519075
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.476466178894043,
"example_count": 600... |
183f5b66-4119-4fce-abde-598f76f5a5bd | easy | jordanrubin | 2026-08-14 01:23:14.100028+00:00 | succeeded | 69ece06bd94846b5a380dc94d4c00c6798b85f8c1e9f0ba91ba31bcdb9376a77 | 29,580 | null | """Weight-tied MLP-cell loop over soft digit states (mlploop).
Cell = MLP over [expected digits, pairwise digit products, RNS residue
simplices (fixed differentiable mixing over Z_p), N digits]. The cell that
learns one-step modular squaring from direct pairs (52% unseen at 14k rows,
day-20 screen) inside the exact-T ... | {
"id": "183f5b66-4119-4fce-abde-598f76f5a5bd",
"created_at": "2026-08-14 01:23:14.100028+00:00",
"db_md5": "0cdede812da1d9957f86c9ea89ed1f0f",
"submitter": "Jordan Rubin",
"github_login": "jordanrubin",
"run_id": "0745e741-251f-4326-83f3-e37358540bc4",
"tier": "easy",
"dataset_id": "e5",
"status": "s... | {
"score": {
"mean_loss": 2.493103733206121,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.060416665238638714
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.160372091649359,
"example_count": 600,... |
1842a056-7286-4e08-9064-236b8c527857 | easy | jordanrubin | 2026-08-06 07:46:44.885887+00:00 | succeeded | 59f3d5359208def9492e2a2accea216538b7fbe379b1df2ed3604a99e6857513 | 15,972 | null | """Gauge-free pair-relational categorical looped Transformer.
This is a learned transition model, not an arithmetic program. The fixed
parts expose only the public prompt grammar and six representation priors:
right-origin decimal slots, a shared ordered digit geometry, unordered pair
tokens, immutable N recall, a so... | {
"id": "1842a056-7286-4e08-9064-236b8c527857",
"created_at": "2026-08-06 07:46:44.885887+00:00",
"db_md5": "89d3c573adc720aa61b1799c6b2dce82",
"submitter": "Jordan Rubin",
"github_login": "jordanrubin",
"run_id": "e71226d6-37fa-4ae7-897b-0d0b2c62db0f",
"tier": "easy",
"dataset_id": "e3",
"status": "s... | {
"score": {
"mean_loss": 2.1827074697067164,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.000625
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.18433968520335,
"example_count": 800,
"... |
1854b175-fbf4-43de-8a0e-03941046a7b7 | easy | datathesisai | 2026-08-05 08:33:25.493772+00:00 | succeeded | 178f35a7c051bc7d739efa6a019a4479a780f1038384dbd2128ee00030435fcf | 20,977 | null | """Causal semigroup tape machine with a learned ordered input router.
The raw sequence is encoded token by token and partitioned by a differentiable
two-change-point router into an immutable program tape, an initial mutable work
tape, and a clock summary. The clock is compressed to one scalar and can only
control a m... | {
"id": "1854b175-fbf4-43de-8a0e-03941046a7b7",
"created_at": "2026-08-05 08:33:25.493772+00:00",
"db_md5": "efeec3079d91d4ece0f14dbc4c175c03",
"submitter": "datathesisai",
"github_login": "datathesisai",
"run_id": "a7920052-a92d-4ed2-bcee-43408d6c06a0",
"tier": "easy",
"dataset_id": "e5",
"status": "... | {
"score": {
"mean_loss": 5.337222099304199,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.012500000186264515
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 5.479938983917236,
"example_count": 600,... |
185a1555-8fd4-4ff4-9e2e-ca58def71426 | easy | DDanlov | 2026-08-14 06:59:46.494514+00:00 | succeeded | 09aa9cb822cf1048e2e61f3e5864e510a4a6b2c1991f427580d3f908c97fa483 | 21,978 | null | import math
from typing import Optional
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import Optimizer
try:
from benchmark import Submission, assert_model_state, OptimizerBundle
except ImportError:
try:
from client import Submission, assert_model_state, OptimizerBu... | {
"id": "185a1555-8fd4-4ff4-9e2e-ca58def71426",
"created_at": "2026-08-14 06:59:46.494514+00:00",
"db_md5": "610645cd550b57af757c9247642c11f8",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "6bb233bd-d8e7-422a-a316-3c1f183855de",
"tier": "easy",
"dataset_id": "e3",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.324221535594397,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.018125000000000002
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.3231789610392917,
"example_count": 800... |
185c74ab-9c4b-4928-95ad-baaaecd162ab | easy | DDanlov | 2026-08-18 08:58:44.502316+00:00 | succeeded | 2daa3dabbf47e1f004651c2e7895944312e775b363dd9a9f753435d4e887457f | 23,511 | null | """
Official Submission for One Layer Deeper Challenge
Architecture: 4-Layer Recurrent Core Pure No-Norm No-Skip DKS Reasoning Model:
- 4 Distinct Pure No-Norm No-Skip Transformer Layers in the Recurrent Core: L1 -> L2 -> L3 -> L4
- Unrolled T times for serial modular squaring
- dim=512, 16 heads, d_ff=1024, RohanShamp... | {
"id": "185c74ab-9c4b-4928-95ad-baaaecd162ab",
"created_at": "2026-08-18 08:58:44.502316+00:00",
"db_md5": "a45a27d6e56da3f75f753d1345b5cf43",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "f94d85c5-b152-4015-9715-3b34f7c4c660",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.4299933665468814,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.05500000149011612
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.368396849774603,
"example_count": 100,... |
186951ac-4b4c-4650-b1f7-d18c6ca08299 | easy | TZZheng | 2026-08-06 21:21:10.181053+00:00 | succeeded | ef4a82d9d5b384d9825eda4333db737ff33346e7bf74adce91734b578d908509 | 7,316 | null | """Self-contained standard Transformer scaling-sweep submission."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
assert_model_state,
)
VARIANT = {'block... | {
"id": "186951ac-4b4c-4650-b1f7-d18c6ca08299",
"created_at": "2026-08-06 21:21:10.181053+00:00",
"db_md5": "b7860fc155dfc9af985d679b586609e1",
"submitter": "TZZheng",
"github_login": "TZZheng",
"run_id": "745a92a0-d835-440e-8fd7-28576f66c983",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 1.872235655784607,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0566666666418314
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.8224111795425415,
"example_count": 100,
... |
187054b6-4241-48ef-b8d4-7108e3aba26a | easy | yashkant | 2026-08-19 02:01:07.216258+00:00 | succeeded | 7224a5041ad2e2cfea4495c1924e13d56ecf3e85840ccf359bb01987efac71ad | 43,124 | null | """Round2231 trains with reset corrections, then ramps to carry.
The correction carry mix is zero through completed update 250, rises
linearly for 300 updates, and is one from update 550 onward. Evaluation
always uses full carry. Architecture, parameters, recurrent work, loss,
optimizer, learning-rate schedule, SAM, a... | {
"id": "187054b6-4241-48ef-b8d4-7108e3aba26a",
"created_at": "2026-08-19 02:01:07.216258+00:00",
"db_md5": "502f75dd1c0e81df2251a485d86b091f",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "bf6be4e7-4f68-4e92-b5b1-ba39fe3a6562",
"tier": "easy",
"dataset_id": "e4",
"status": "succeed... | {
"score": {
"mean_loss": 2.070107940626687,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0104166666790843
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.0730891615582987,
"example_count": 1200,... |
1873916d-8455-450e-8c84-ab860eba60a6 | easy | yashkant | 2026-08-11 03:17:38.297071+00:00 | succeeded | 99a57f483913b01ff4d510949e4ffe1a41774c166d40dfb78b9458d12c600057 | 35,176 | null | """Round1377 compact extragradient TRM and committed-step control.
A width-32 answer/latent pair is refined by one shared two-layer MLP. Every
public-T macro first predicts a provisional pair, reevaluates the same MLP at
that lookahead point, and either applies the lookahead residual to the
original pair (candidate) ... | {
"id": "1873916d-8455-450e-8c84-ab860eba60a6",
"created_at": "2026-08-11 03:17:38.297071+00:00",
"db_md5": "438eb17c22333cfc7991f7776797f9a2",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "c6b38ab7-f266-4876-98f7-8dca7825b6dd",
"tier": "easy",
"dataset_id": "e2",
"status": "succeed... | {
"score": {
"mean_loss": 2.482729196548462,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.04520833492279053
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.824861764907837,
"example_count": 300,
... |
1873c426-2b05-409a-ac19-d882afff80af | easy | DDanlov | 2026-08-21 02:21:48.604129+00:00 | failed | 50821d60df1918962985d7131f4f3bf20890e6d51c7262f18587f1927723e94e | 25,133 | null | from __future__ import annotations
import math
import time
from typing import Any, Optional, Tuple
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import Optimizer
try:
from benchmark.api import ModelSpec, OptimizerBundle, OptimizerSpec, Submission, assert_model_state
except Imp... | {
"id": "1873c426-2b05-409a-ac19-d882afff80af",
"created_at": "2026-08-21 02:21:48.604129+00:00",
"db_md5": "0a319749c0ce4cec482dd43c29a77e10",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "7cdc3e60-db78-4f97-a01a-78480fa50f00",
"tier": "easy",
"dataset_id": "e5",
"status": "failed",
... | null
|
1876f1ab-484d-4ba4-a2ee-b04d1aabe94d | easy | 0xDEADD06 | 2026-08-12 13:31:56.230417+00:00 | succeeded | f8ebb3070a558fac3db400875d6232bc23377158882b9856bd32a50f3c623ea5 | 5,900 | null | """ParaLSTM: recurrent LSTM with weight-tied passes for algorithmic tasks."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import nn, Tensor
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
assert_model_state,
)
D_MODEL... | {
"id": "1876f1ab-484d-4ba4-a2ee-b04d1aabe94d",
"created_at": "2026-08-12 13:31:56.230417+00:00",
"db_md5": "984d66ba22f2b60aa0d2cd05cccabb64",
"submitter": "0xDEADD06",
"github_login": "0xDEADD06",
"run_id": "e7ec2ee9-794c-4b53-b9a0-b4efe5b19309",
"tier": "easy",
"dataset_id": "e5",
"status": "succee... | {
"score": {
"mean_loss": 2.3496094209661664,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.005416666828095913
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.3490017112595085,
"example_count": 60... |
1879ce83-ee06-4253-9ac1-0b441cfaafa8 | easy | khushidahi | 2026-08-20 03:47:27.601585+00:00 | succeeded | c77626ecdbf57467486307492031a116e10d6c7c8ed64fe283eeb1f37c654ae6 | 33,719 | null | """R297: semantic-factorized tied causal functional root.
This candidate is deliberately diagnostic. It trains on evaluator-owned final
labels at several depths so that one shared transition must compose. The model
never converts N or X to native numbers and never implements multiplication or
modular reduction.
The... | {
"id": "1879ce83-ee06-4253-9ac1-0b441cfaafa8",
"created_at": "2026-08-20 03:47:27.601585+00:00",
"db_md5": "45accd4454e0743bde578aa75506866e",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "34b094e9-8433-4ab7-9a18-9b04923c8fa7",
"tier": "easy",
"dataset_id": "e3",
"status": "succ... | {
"score": {
"mean_loss": 2.2672119582534003,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.005000000009313226
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2557271716471323,
"example_count": 80... |
188335a7-cedf-4e05-8728-53ea2bbc323d | easy | erdavis0 | 2026-08-22 02:35:40.794171+00:00 | succeeded | 23b0780f2d067145948e433ff7b2bf5913b31bd724a5c1ac237c39eb1c08b842 | 6,295 | null | """Additive matched control for the prefix-suffix moment mixer."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
assert_model_state,
)
WIDTH = 64
HIDDEN ... | {
"id": "188335a7-cedf-4e05-8728-53ea2bbc323d",
"created_at": "2026-08-22 02:35:40.794171+00:00",
"db_md5": "6fa839cd6fd1aedca88beef10107540d",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "19355bbb-eae0-4703-9fda-9c64ea2a3bcc",
"tier": "easy",
"dataset_id": "e3",
"status": "succeeded",... | {
"score": {
"mean_loss": 7.1689043045043945,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.008124999701976776
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 7.411953449249268,
"example_count": 800... |
188369a1-6da2-434e-9e10-1a3862d7fcfb | easy | chuk-uzowihe | 2026-08-06 23:44:10.724313+00:00 | succeeded | 9f4b45f7fc79854faa2831968f170822757161a76cc07825a7c5c7202dfd59ed | 33,677 | null | """One Layer Deeper submission: FPRM over a digit lattice with a Fourier T latent.
Structure (v4)
--------------
1. N and x are *digit slots*: each digit is gathered into a dense per-number
grid (column j = place value 10^j, short rows left-padded with explicit
zero digits) and embedded as digit token + learned ... | {
"id": "188369a1-6da2-434e-9e10-1a3862d7fcfb",
"created_at": "2026-08-06 23:44:10.724313+00:00",
"db_md5": "540435dd4f5ac4f3225599aabafbe7d5",
"submitter": "Chuk Uzowihe",
"github_login": "chuk-uzowihe",
"run_id": "ee3e6244-07ca-475d-ba26-0da91721b51a",
"tier": "easy",
"dataset_id": "e4",
"status": "... | {
"score": {
"mean_loss": 13.681152664152368,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.007708333358168602
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 13.812199108625563,
"example_count": 12... |
1892e3a9-a9c0-4d9b-a1ed-db2ed7af210b | easy | shreyash-chonkie | 2026-08-24 17:42:59.130928+00:00 | succeeded | 186e3e08d6b15d08f89242d27fc941fca95fa77feefccef8fae8e1f267178426 | 10,732 | null | """Gated recurrent attention trained from terminal cross-entropy."""
from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
assert_... | {
"id": "1892e3a9-a9c0-4d9b-a1ed-db2ed7af210b",
"created_at": "2026-08-24 17:42:59.130928+00:00",
"db_md5": "3345422be8fbbfa4ce7b98ac44d10219",
"submitter": "Shreyash",
"github_login": "shreyash-chonkie",
"run_id": "30cf6096-f4f8-43bb-91f7-d3510f3ac80a",
"tier": "easy",
"dataset_id": "e6",
"status": "... | {
"score": {
"mean_loss": 9.240381717681885,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.026801803149282932
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 9.643109321594238,
"example_count": 60,
... |
189590ac-13af-450c-9209-ee359c293c35 | easy | jordanrubin | 2026-08-11 01:38:28.201158+00:00 | succeeded | aa0c446a01b1c50a3739ac906b125fbe05d6b1cd9cf0a23d1a564e062285e8d6 | 21,670 | null | """Weight-tied MLP-cell loop over soft digit states (mlploop).
Cell = MLP over [expected digits, pairwise digit products, RNS residue
simplices (fixed differentiable mixing over Z_p), N digits]. The cell that
learns one-step modular squaring from direct pairs (52% unseen at 14k rows,
day-20 screen) inside the exact-T ... | {
"id": "189590ac-13af-450c-9209-ee359c293c35",
"created_at": "2026-08-11 01:38:28.201158+00:00",
"db_md5": "344e901468d7fc4aff5006fca563aa6e",
"submitter": "Jordan Rubin",
"github_login": "jordanrubin",
"run_id": "98befcde-e3eb-461a-aa2d-a947d394fa8c",
"tier": "easy",
"dataset_id": "e5",
"status": "s... | {
"score": {
"mean_loss": 2.443575113587208,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.009166666461775701
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1562784262034924,
"example_count": 600... |
189913b1-94c7-46d7-a7b9-465a7c322ec2 | easy | DDanlov | 2026-08-14 06:45:16.252674+00:00 | succeeded | 26b251cc3b81d81a536f8f36a9cec059412b13ec89e54284a3113ef5cc8640ca | 21,978 | null | import math
from typing import Optional
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import Optimizer
try:
from benchmark import Submission, assert_model_state, OptimizerBundle
except ImportError:
try:
from client import Submission, assert_model_state, OptimizerBu... | {
"id": "189913b1-94c7-46d7-a7b9-465a7c322ec2",
"created_at": "2026-08-14 06:45:16.252674+00:00",
"db_md5": "29f8fe5a56a43b429b624eae20c07b04",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "41454ecc-d32a-4fef-a237-907b8caeab17",
"tier": "easy",
"dataset_id": "e3",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.833183334464667,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.018125000000000002
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.8331833020147275,
"example_count": 800... |
18a1893c-8b0f-4877-9c22-9f32670b805f | easy | oupadhyay | 2026-08-29 18:44:27.187849+00:00 | succeeded | 7298226707ec378fd3c07cb39b600dda5b3c108712d76e19937ad46d484b9387 | 3,993 | null | """Conditional optimizer phase 3: warm8_wd0."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import nn
from benchmark import ModelSpec,OptimizerSpec,OptimizerBundle,Submission,assert_model_state
W,D,H=4,64,32
class C:
def __init__(s,vocab_size,max_seq_len):s.vocab_... | {
"id": "18a1893c-8b0f-4877-9c22-9f32670b805f",
"created_at": "2026-08-29 18:44:27.187849+00:00",
"db_md5": "a8eac7c50f66498494c6b306b5279403",
"submitter": "Ojasw Upadhyay",
"github_login": "oupadhyay",
"run_id": "29df6e76-4999-4461-aa22-3e7e96b43c09",
"tier": "easy",
"dataset_id": "e6",
"status": "s... | {
"score": {
"mean_loss": 2.0121511220932007,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.03536036144942045
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.0128226280212402,
"example_count": 60,... |
18a8775c-1a46-4cf5-9eae-2bf7d581a7db | easy | Yalyenea | 2026-08-12 13:04:46.806218+00:00 | succeeded | 1b2d3737aa2b0dd5d6dac18cf443ec59ebe87c38e89db25ad4b48f30c5f5235f | 11,733 | null | """Class-gated dropout recurrence with 8x T=1 weighting and WSD."""
from __future__ import annotations
import math
import time
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
... | {
"id": "18a8775c-1a46-4cf5-9eae-2bf7d581a7db",
"created_at": "2026-08-12 13:04:46.806218+00:00",
"db_md5": "3412ced66080f2b805ed54440f41aacf",
"submitter": "yfff",
"github_login": "Yalyenea",
"run_id": "3ee666f0-5b17-4276-a369-e953ab111c85",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded",
... | {
"score": {
"mean_loss": 3.2038971067008584,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.011250000000000001
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2817677350322225,
"example_count": 60... |
18aee6d2-fab3-4ec8-924e-cc7fe2d84bb4 | easy | Dandandan | 2026-08-07 02:55:50.295799+00:00 | failed | 8d212f6334e71bb818f1deb1b79aa0398aaeed4ffb67dea404ceed16e2646e74 | 98,978 | null | """Generalized exact recursive learner for One Layer Deeper.
The learned transition is a population of ordinary modular neural units with
several generic arithmetic activations. A single learned evidence selector
chooses one transition, which is then reused for every input-requested outer
step. The population covers ... | {
"id": "18aee6d2-fab3-4ec8-924e-cc7fe2d84bb4",
"created_at": "2026-08-07 02:55:50.295799+00:00",
"db_md5": "77aa66f5b55752d3e7229ed48123f71b",
"submitter": "Daniël Heres",
"github_login": "Dandandan",
"run_id": "34c46e46-ecb4-4a23-a067-ffa7bf7d89be",
"tier": "easy",
"dataset_id": "e1",
"status": "fai... | null
|
18b0ee4b-c8f8-41b7-8a60-aae8ac5bc9b7 | easy | shirvani-jr | 2026-08-29 01:21:19.690320+00:00 | succeeded | 2576c87c3f7488e3d4c646e2c6fa7cac5f884e2775bd5719b057b93833f5d4e8 | 50,077 | null | """Adaptive learned discrete recurrent fabric for One Layer Deeper.
A gradient-trained, operation-free recurrent computational machine. The
model learns, from the competition's endpoint supervision: how to parse the
prompt into a bank of categorical digit registers, a reusable transition
circuit applied serially, a l... | {
"id": "18b0ee4b-c8f8-41b7-8a60-aae8ac5bc9b7",
"created_at": "2026-08-29 01:21:19.690320+00:00",
"db_md5": "468e24784c5e4bd221954034dd34758c",
"submitter": "Ali",
"github_login": "shirvani-jr",
"run_id": "b0d902ef-2e0e-4b7c-9b1d-e2ac4f5d3915",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 0.7977999001741409,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.8133333325386047
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.45201411843299866,
"example_count": 100... |
18c17ba6-c141-4454-8ece-917b063eb998 | easy | anirudh-chakravarthy | 2026-08-12 03:26:57.387430+00:00 | succeeded | abc0b3fa0f542ba960e2941f45ef63ecec31e6c3ba2a492af89d35d6d21ced65 | 14,572 | null | """A learned state transition applied once per requested time step."""
from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
asser... | {
"id": "18c17ba6-c141-4454-8ece-917b063eb998",
"created_at": "2026-08-12 03:26:57.387430+00:00",
"db_md5": "c2569ee829d66479bef77c48c3e5b6ed",
"submitter": "Anirudh S Chakravarthy",
"github_login": "anirudh-chakravarthy",
"run_id": "d3c15ae4-7a88-4126-983b-948acebc3455",
"tier": "easy",
"dataset_id": "... | {
"score": {
"mean_loss": 3.9448776274903103,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.013750000018626453
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.396921412604769,
"example_count": 800... |
18c878c5-f623-43ab-ad4c-8ff5572be9b6 | easy | sapient-sapiens | 2026-08-17 19:45:25.009758+00:00 | succeeded | 73b24d001589dea7b46cf54d4671cee005cbfcf7bd92b0eb9133263887ecc018 | 28,660 | null | """v40 lineage: T-free operator, N memory, PonderNet mix, no T in the loss.
Field-local encoder, T-free core, countable halt, answer-slot readout.
N is seeded and restored every step. T is never parsed in forward or loss.
"""
from __future__ import annotations
import math
import time
import torch
import torch.nn.fu... | {
"id": "18c878c5-f623-43ab-ad4c-8ff5572be9b6",
"created_at": "2026-08-17 19:45:25.009758+00:00",
"db_md5": "ab0995ba57df7797ce934d12bf6df7e5",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "e6a13a1d-90f7-44fa-9ec9-a1c9205113fd",
"tier": "easy",
"dataset_id": "e5",
"status": "su... | {
"score": {
"mean_loss": 2.1458609242246425,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.009583333345750969
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.149260156491412,
"example_count": 600... |
18d13cad-e99a-43ec-b76c-67bc5ecd8239 | easy | alirezashirvani-jr | 2026-08-29 12:51:36.283683+00:00 | succeeded | eb6c1a60bb5b69554a6c86362d97538badbe469fdbfca1936d94a128a36f76a4 | 32,602 | null | from __future__ import annotations
import math
import time
from contextlib import nullcontext
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
assert_model_state,
)
MODEL_... | {
"id": "18d13cad-e99a-43ec-b76c-67bc5ecd8239",
"created_at": "2026-08-29 12:51:36.283683+00:00",
"db_md5": "16669e8364313e16bc8075816a4226a4",
"submitter": "alirezashirvani-jr",
"github_login": "alirezashirvani-jr",
"run_id": "00e3c78a-c9f1-428b-a9a0-bebe5672b175",
"tier": "easy",
"dataset_id": "e5",
... | {
"score": {
"mean_loss": 3.065808031338829,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.04541666675359011
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.0909125804901123,
"example_count": 600,... |
18d788fa-100a-4662-8e4b-5690188e2487 | easy | DDanlov | 2026-08-20 11:44:40.120376+00:00 | succeeded | 586745384c4fa422a00dcb8e96438585edc7a20b129982741361f1a9ec7f4e93 | 20,911 | null | from __future__ import annotations
import math
import time
from typing import Any, Optional, Tuple
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import Optimizer
try:
from benchmark.api import ModelSpec, OptimizerBundle, OptimizerSpec, Submission, assert_model_state
except Imp... | {
"id": "18d788fa-100a-4662-8e4b-5690188e2487",
"created_at": "2026-08-20 11:44:40.120376+00:00",
"db_md5": "1edbcfbfc146057514cdc75a9591a574",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "f0215f18-7d2d-4df7-9574-87df9d5eed7a",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.29064873052448,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.014583333345750966
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2872743595875966,
"example_count": 600,... |
18e15bd1-08a7-4f92-bee7-5fdfa378f278 | easy | erdavis0 | 2026-08-26 03:28:22.698900+00:00 | succeeded | 6fe6fdf46bcdc9c76ffef64fbdc6b302f2262f8879ac4d7daea4e62dccf8b390 | 30,354 | null | """Latent-role continuous REV C5: hybrid token CE and sequence mixture.
Public delimiter syntax only routes opaque token categories into distinct,
right-aligned N, X, and T tapes. Learned local cells compile separate
interaction, carry, and modulus-conditioned workspaces. N is isolated from
interaction and carry and... | {
"id": "18e15bd1-08a7-4f92-bee7-5fdfa378f278",
"created_at": "2026-08-26 03:28:22.698900+00:00",
"db_md5": "14ff631f83062f36483f24b1bd94c4c9",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "865b88fd-459a-46a9-932c-c9543ee1249d",
"tier": "easy",
"dataset_id": "e3",
"status": "succeeded",... | {
"score": {
"mean_loss": 6.297380983710594,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.011875000009313226
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 6.389037201672378,
"example_count": 800,... |
18e4605e-df5b-417b-9d84-5ac2f50076a5 | easy | DDanlov | 2026-08-11 22:26:49.199212+00:00 | succeeded | cb82f4612f44bf9156e4b5c494814bfc2d10b40afb1fb6be8f6d595ec45220f3 | 20,818 | null | from __future__ import annotations
import sys
import os
import math
import time
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim.optimizer import Optimizer
VOCAB_SIZE = 17
DIGIT_OFFSET = 7
def init_cutting_the_skip(module):
"""SVD-based orthogonal initialization to ensure initia... | {
"id": "18e4605e-df5b-417b-9d84-5ac2f50076a5",
"created_at": "2026-08-11 22:26:49.199212+00:00",
"db_md5": "320221ceccae83253f74dfe4afba5138",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "493037c4-9dad-42ba-b411-fd27c915ec4c",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 266.12511698047604,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0004166666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 302.0178305279514,
"example_count": 60... |
18f29977-b290-4269-a3f3-755ed3aa8487 | easy | ehonig | 2026-08-20 18:12:34.320384+00:00 | succeeded | f81637fd0030281cc4e43524d2bfbeceab8a861f930c55ae6a6711c6f0e19a28 | 25,078 | null | """A plain Transformer for repeated modular squaring, and nothing else yet.
This is a deliberate restart. The previous submission accumulated a field
parser, a carry scan, a learned reciprocal, periodic features, a digit
bottleneck, a depth selector and two cell types, most of them tuned around a
memorisation table th... | {
"id": "18f29977-b290-4269-a3f3-755ed3aa8487",
"created_at": "2026-08-20 18:12:34.320384+00:00",
"db_md5": "da66c90ca5f3d7275f3ac07141bf1e6d",
"submitter": "Edouardo Honig",
"github_login": "ehonig",
"run_id": "d72c8612-0c80-4e62-9e7f-94923fc264ef",
"tier": "easy",
"dataset_id": "e1",
"status": "succ... | {
"score": {
"mean_loss": 2.144800066947937,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.06500000134110451
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.159137725830078,
"example_count": 100,
... |
18f65c02-3858-4676-8475-55a33719887f | easy | DDanlov | 2026-08-15 21:05:18.222384+00:00 | succeeded | 3eb517cb9667863720ce75d439fd67ebc36b79a26534e0b145dad12284313550 | 17,950 | null | """
Official Submission for One Layer Deeper Challenge
Architecture: Deep Equilibrium Port-Hamiltonian Deliberation Network (Model C v3 / DEPH-Net)
- Input: Multi-Scale Linear Harmonic Torus Frequency Encoder (omega_j = 1..64)
- Recurrent Core: Continuous Port-Hamiltonian Dynamical Flow dz/dt = (J - R) * grad_E
* J =... | {
"id": "18f65c02-3858-4676-8475-55a33719887f",
"created_at": "2026-08-15 21:05:18.222384+00:00",
"db_md5": "623970f9d216ffb45b2ef4f8507de941",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "8503caaa-2115-4f8c-bd40-272e97dfeb30",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.4699483689325428,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.002916666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.4704458578552364,
"example_count": 60... |
1901fe86-773e-42b6-9259-5c4967ee11bf | easy | sankalp1999 | 2026-08-04 12:14:43.454878+00:00 | succeeded | 264ac27a3b6e775e68546dea0b66d8548f42abf48a5aedd3c7c419dd4c9fb499 | 7,511 | null | """Latent recurrence: encode once, apply a shared squaring map T times, decode.
v4 put the loop inside a transformer and underfit badly (train acc 4% after
4,432 steps) because attention cost ~7ms/step at 13 tokens and bought nothing.
v2 fit the train set perfectly with a cheap flattened encoder but memorized,
because... | {
"id": "1901fe86-773e-42b6-9259-5c4967ee11bf",
"created_at": "2026-08-04 12:14:43.454878+00:00",
"db_md5": "65280dd820afb3c3ab0ae45e61454099",
"submitter": "Sankalp ",
"github_login": "sankalp1999",
"run_id": "81daef20-342c-4664-932f-440a3a7a35fd",
"tier": "easy",
"dataset_id": "e5",
"status": "succe... | {
"score": {
"mean_loss": 2.1603285946911814,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.002916666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1640015503750787,
"example_count": 60... |
19085405-9f10-45ce-9b7b-0e946a73d503 | easy | Dandandan | 2026-08-05 22:26:53.402864+00:00 | succeeded | cf63a5cf6a2f46bb38abff9e855f809b23cb30f1ed38de4b537f9579a8903f40 | 41,151 | null | """End-to-end recursive function learner for One Layer Deeper.
A population of randomly initialized polynomial register networks is reused
exactly once per requested outer step. The evaluator trains every coefficient
and the evidence selector through its normal forward/backward/optimizer loop.
There is no recurrence-... | {
"id": "19085405-9f10-45ce-9b7b-0e946a73d503",
"created_at": "2026-08-05 22:26:53.402864+00:00",
"db_md5": "67701dfb09c9f4d71ad540e59bfed9f1",
"submitter": "Daniël Heres",
"github_login": "Dandandan",
"run_id": "fcab6e66-e1a3-45db-8c47-1e3890f551b2",
"tier": "easy",
"dataset_id": "e5",
"status": "suc... | {
"score": {
"mean_loss": 1.1348570754328067e-05,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.1365176222235958e-05,
"example_count": 600,
... |
190eb6b4-31e7-4a31-87f1-0d598ee1a4d5 | easy | Yalyenea | 2026-08-11 21:13:06.622891+00:00 | succeeded | e0fc33e684ec8aa5b7908dba1f7a3acd7444ca5c247a2c8ce652542b450f9751 | 12,305 | null | """R4 full-bandwidth candidate with 0.02 carried-state jitter and cosine horizon 2000."""
from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
assert... | {
"id": "190eb6b4-31e7-4a31-87f1-0d598ee1a4d5",
"created_at": "2026-08-11 21:13:06.622891+00:00",
"db_md5": "e3a9c6c31ed1f446b943daeee479dc85",
"submitter": "yfff",
"github_login": "Yalyenea",
"run_id": "8875e8c4-3b0d-4fab-a127-c35e94c08f91",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded",
... | {
"score": {
"mean_loss": 6.358806570639482,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.014999999897554517
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 6.224160671234131,
"example_count": 600,... |
19130dc8-0309-4a03-bd52-5f179f898725 | easy | arseniivanov | 2026-08-10 02:19:42.904368+00:00 | succeeded | d0d6815201e92e2925afa513b9c760d0db189ac73f69a900ff7916255340ad35 | 21,875 | null | """Looped, vector-quantized, adaptively-halting Transformer.
Depth is created by iteration over a weight-tied step operator, not by stacking layers:
encoder tokens -> N-context untied, read once; hoists the modulus-dependent
work out of the recurrence
st... | {
"id": "19130dc8-0309-4a03-bd52-5f179f898725",
"created_at": "2026-08-10 02:19:42.904368+00:00",
"db_md5": "183065617c572ea7d58c087d3a786d68",
"submitter": "Arseni Ivanov",
"github_login": "arseniivanov",
"run_id": "dbe6587d-aeff-4c7b-ac65-8f4aea13a111",
"tier": "easy",
"dataset_id": "e1",
"status": ... | {
"score": {
"mean_loss": 21.180971145629883,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 19.57587242126465,
"example_count": 100,
"exac... |
1919b6ee-27b3-434e-9edf-0877e64dab15 | easy | khushidahi | 2026-08-16 11:55:08.905783+00:00 | succeeded | 711d51257a1ad8f85efca63e6ab59274db8deffacc124fb18b6585ec87f54692 | 24,102 | null | """R92: ordered-modulus transition with global tape mixing.
This candidate is deliberately diagnostic. It trains on evaluator-owned final
labels at several depths so that one shared transition must compose. The model
never converts N or X to native numbers and never implements multiplication or
modular reduction.
T... | {
"id": "1919b6ee-27b3-434e-9edf-0877e64dab15",
"created_at": "2026-08-16 11:55:08.905783+00:00",
"db_md5": "388d154c1dd73c70420565e44bb47c69",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "3f642e0c-7b82-4142-971e-d9c04d5d9d33",
"tier": "easy",
"dataset_id": "e5",
"status": "succ... | {
"score": {
"mean_loss": 2.54614135548662,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.03375
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.590765716783669,
"example_count": 600,
"ex... |
191e18f6-a48f-4b7a-8a89-20358620f2bd | easy | ddoan | 2026-08-14 13:20:40.581646+00:00 | succeeded | 5ea9968dfd6066da5871262814108e0e25104078557b783cc4543c68482d034d | 9,822 | null | """Tiny continuous place tape with an N-conditioned quadratic fold."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
BatchReuseContext,
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,... | {
"id": "191e18f6-a48f-4b7a-8a89-20358620f2bd",
"created_at": "2026-08-14 13:20:40.581646+00:00",
"db_md5": "faf7e5db75a505c4421525f1953a6821",
"submitter": "Doug Doan",
"github_login": "ddoan",
"run_id": "4a30731c-dc2f-48a2-8a3d-dfe3779d3a00",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 3.5519015789031982,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.03499999828636646
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 5.022947311401367,
"example_count": 100,... |
19226c29-6ab9-4123-8eee-b4003531cd1d | easy | DDanlov | 2026-08-10 09:34:44.085112+00:00 | succeeded | e83c03fc8f5eb9fd5c4cb731d3f22b61a926795de4a084f1b3081f93259c74af | 15,745 | null | from __future__ import annotations
import sys
import os
import math
import time
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim.optimizer import Optimizer
VOCAB_SIZE = 17
DIGIT_OFFSET = 7
def init_cutting_the_skip(module):
if isinstance(module, nn.Linear):
nn.init.ortho... | {
"id": "19226c29-6ab9-4123-8eee-b4003531cd1d",
"created_at": "2026-08-10 09:34:44.085112+00:00",
"db_md5": "03573171215fce109a1a9d3924549fba",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "c5ce41cb-b055-4cb9-93fa-4f6cd690bdf9",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.4758815604849644,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0833333307504654
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.4966659545898438,
"example_count": 100,... |
1924cd69-8fc5-418f-9309-5b77f933da56 | easy | jordanrubin | 2026-08-12 04:53:02.585334+00:00 | succeeded | a8113fa5fd30cd348b465897ea29b52cff29e3c0818fa036daff3b32834a3bbe | 22,495 | null | """Weight-tied MLP-cell loop over soft digit states (mlploop).
Cell = MLP over [expected digits, pairwise digit products, RNS residue
simplices (fixed differentiable mixing over Z_p), N digits]. The cell that
learns one-step modular squaring from direct pairs (52% unseen at 14k rows,
day-20 screen) inside the exact-T ... | {
"id": "1924cd69-8fc5-418f-9309-5b77f933da56",
"created_at": "2026-08-12 04:53:02.585334+00:00",
"db_md5": "73f7cb6eef76bd925bbd37d57d0aa13b",
"submitter": "Jordan Rubin",
"github_login": "jordanrubin",
"run_id": "f35ed353-85a8-4985-8eab-fe7d668267aa",
"tier": "easy",
"dataset_id": "e5",
"status": "s... | {
"score": {
"mean_loss": 2.6249286986865483,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.012499999642993014
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1732071618060895,
"example_count": 60... |
193eb191-7a47-452e-bfbe-8cef741c098f | easy | yashkant | 2026-08-13 00:24:18.829711+00:00 | succeeded | 1785fc0c1a760aee525a8560bb9a2df9a9b91dd8304eefc9dd647d8cfa3994f3 | 28,598 | null | """Round1174 equal-work recurrence-attachment-disabled control."""
from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
assert_model_state,
)
from torch import... | {
"id": "193eb191-7a47-452e-bfbe-8cef741c098f",
"created_at": "2026-08-13 00:24:18.829711+00:00",
"db_md5": "a8202c39f1d63bfda0273582016519a4",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "2e94d2d5-0535-4977-91cc-496f19c44aa6",
"tier": "easy",
"dataset_id": "e2",
"status": "succeed... | {
"score": {
"mean_loss": 4.085277318954468,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.3200000040233135
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 6.385202407836914,
"example_count": 300,
... |
19495c41-3b85-4cc1-8a6e-2044f0ead2fe | easy | chad-atexpedient | 2026-08-20 16:22:37.415933+00:00 | succeeded | db7975c80ec1e7d9f860b7de0c5f88252eb726f73dfbb24df58fb316b8d98c30 | 22,233 | null | """X124: strict-parity routed-feature ablation.
Before building further on X98E, test whether its routed pair lane causally helps.
X124 changes only the existing routed scale to a fixed zero contribution while
keeping every parameter, shape, and RNG draw. Pair features are still computed,
but cannot affect the transit... | {
"id": "19495c41-3b85-4cc1-8a6e-2044f0ead2fe",
"created_at": "2026-08-20 16:22:37.415933+00:00",
"db_md5": "10cae3a4167619adad1e0a7bab0e8983",
"submitter": "chad-atexpedient",
"github_login": "chad-atexpedient",
"run_id": "e599f20f-5296-4cd9-8899-11b4c1204281",
"tier": "easy",
"dataset_id": "e5",
"st... | {
"score": {
"mean_loss": 2.308705890690044,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.007500000037252903
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.332973262119721,
"example_count": 600,... |
194b2b69-af70-4bce-8b86-70605df3daca | easy | jayyala | 2026-08-16 23:06:36.324028+00:00 | succeeded | 9fb85c86563736265af0d6681567d86fe953dabca7a4c104c60db37f4450ec4e | 14,657 | null | """A scalable recurrent machine with canonical categorical byte states.
Each integer is represented by a short little-endian tape of 256-way slots.
One tied bidirectional mixing cell learns the unknown state transition from the
current tape and the prompt context, then its categorical output is fed back
for exactly T ... | {
"id": "194b2b69-af70-4bce-8b86-70605df3daca",
"created_at": "2026-08-16 23:06:36.324028+00:00",
"db_md5": "abe23542b6010fc06d199e62298e64a8",
"submitter": "jayyala",
"github_login": "jayyala",
"run_id": "23058dde-2ac7-47e1-9b74-7e44e5ff6286",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 34.56010291758958,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.038749999925494195
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 35.313899993896484,
"example_count": 600... |
194eed4d-bc45-4633-9401-18109fca59f0 | easy | alirezashirvani-jr | 2026-08-31 07:30:51.693319+00:00 | succeeded | c252ce15d535434ae9695254770a32b96e97a7120717f3216a16792d4ff23118 | 34,903 | null | from __future__ import annotations
import math
import time
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
assert_model_state,
)
TRACK_FLOOR = 3
TRACK_CEILING = 96
STAGE_B... | {
"id": "194eed4d-bc45-4633-9401-18109fca59f0",
"created_at": "2026-08-31 07:30:51.693319+00:00",
"db_md5": "b9699e2f9827d45ad6df7d21eed39459",
"submitter": "alirezashirvani-jr",
"github_login": "alirezashirvani-jr",
"run_id": "c6b6fb9f-2061-4992-85d3-5bf0235dfcaf",
"tier": "easy",
"dataset_id": "e6",
... | {
"score": {
"mean_loss": 0.01942709181457758,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.0199933722615242,
"example_count": 60,
"exa... |
1950605d-2513-4c8f-9a2c-08ea906b81dd | easy | khushidahi | 2026-08-21 21:12:57.505911+00:00 | succeeded | 59cfb45a802bc55dfb2dce353ebfddb006a442e30bb4938e52bd71d1e168f64d | 33,238 | null | """R366: Balanced Gradient Filtering on the fixed-loop TRM control.
The forward graph and final-label loss are exactly R321. The only mechanism
change is an equal raw/EMA norm-balanced gradient mixture before AdamW. The
EMA is registered optimizer state, so the evaluator counts, device-checks,
and serializes it. Th... | {
"id": "1950605d-2513-4c8f-9a2c-08ea906b81dd",
"created_at": "2026-08-21 21:12:57.505911+00:00",
"db_md5": "edf5833d6a034110a466d0aa59533461",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "1521323f-b889-4e80-8ee4-3b0b368b418f",
"tier": "easy",
"dataset_id": "e4",
"status": "succ... | {
"score": {
"mean_loss": 2.1690536577603803,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0036458333333333334
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1669252983182594,
"example_count": 1... |
195245df-f2df-4fba-93ed-1c6c688a9efe | easy | lzy54 | 2026-08-26 18:04:30.380551+00:00 | succeeded | a2bf9db100295d92759356264796aba3d2bbfdff5550bce42aec6080e9eb6955 | 4,433 | null | """Four-pass shared Transformer with learned concat-adapter input recall."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
assert_model_state,
)
D_MODEL ... | {
"id": "195245df-f2df-4fba-93ed-1c6c688a9efe",
"created_at": "2026-08-26 18:04:30.380551+00:00",
"db_md5": "0c8bd06487ab49be6a002735cc2a06b5",
"submitter": "lzy54",
"github_login": "lzy54",
"run_id": "88d16cb8-bd15-43ee-b409-2d21b38efd35",
"tier": "easy",
"dataset_id": "e6",
"status": "succeeded",
... | {
"score": {
"mean_loss": 5.268638849258423,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.015090090688318014
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 5.066028594970703,
"example_count": 60,
... |
195386ee-27b1-44fc-a94a-db991c93e1c7 | easy | viridale | 2026-08-16 03:58:52.888900+00:00 | succeeded | 19d07d9030dcba5927ae45e13f087d19935191a93e686c7680777deaeed611fd | 18,061 | null | """Integer-polynomial boundary quotient with clip-invariant acquisition.
hypothesis: quotienting the T1 coordinate fiber to one autonomous polynomial neuron
and projecting its learned coefficients to integers makes the acquired boundary
compose exactly and transfer across modulus identities.
axis: arch
target:... | {
"id": "195386ee-27b1-44fc-a94a-db991c93e1c7",
"created_at": "2026-08-16 03:58:52.888900+00:00",
"db_md5": "aa64c42e392b1cd8e58ee153383a4ab4",
"submitter": "priormancer",
"github_login": "viridale",
"run_id": "29990c1d-ee8b-486d-a185-f449c6118dcd",
"tier": "easy",
"dataset_id": "e2",
"status": "succe... | {
"score": {
"mean_loss": 0.1493113711476326,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.146402969956398,
"example_count": 300,
"exac... |
195a7adf-a33c-440d-814d-b7a25683ff5c | easy | alexkreimer | 2026-08-22 07:00:56.698169+00:00 | succeeded | 094f2102e05e6d0b466a4688878afb44f133f490c00db8295499344acd71aaa2 | 10,044 | null | """Tied-recursive latent-depth model, iterated exactly T times per example.
Starting point for One Layer Deeper: one small attention+MLP cell is
applied repeatedly to a shared latent state (weight-tied recursion), with
the original token/position embeddings re-injected at every iteration so
the operands (N, x, T) don'... | {
"id": "195a7adf-a33c-440d-814d-b7a25683ff5c",
"created_at": "2026-08-22 07:00:56.698169+00:00",
"db_md5": "0f517d90509ff8a1e6b758c7c0630c77",
"submitter": "Alex Kreimer",
"github_login": "alexkreimer",
"run_id": "f669a67c-4915-414b-b702-dae437f1f6fa",
"tier": "easy",
"dataset_id": "e1",
"status": "s... | {
"score": {
"mean_loss": 3.4184119712988297,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.02166666674117247
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.6105525724923435,
"example_count": 100... |
195be7e0-9899-4fe8-9565-8e5c9bc23074 | easy | jayyala | 2026-08-18 13:17:27.399395+00:00 | succeeded | e6c747c86aa58869f964c195b1efe51c52ccb8bbc525e3613aff1138a5113d25 | 24,443 | null | """One N-equivariant step: wrap a learned bit-pair product, then invert.
exp(2*pi*i*k/N) depends only on k mod N, so the integer product wraps into
the reduced residue. That wrapping is the transition, reused T times. A
tied coarse-to-fine decoder inverts the phase; it is not a table of (N, x).
Prompt reconstruction... | {
"id": "195be7e0-9899-4fe8-9565-8e5c9bc23074",
"created_at": "2026-08-18 13:17:27.399395+00:00",
"db_md5": "4da96419acfd905f77442ec1acb5a1d0",
"submitter": "jayyala",
"github_login": "jayyala",
"run_id": "3126cc13-0a04-462a-b54a-4432aa51ddf5",
"tier": "easy",
"dataset_id": "e6",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.195112943649292,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.944819837808609
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.5135135650634766,
"example_count": 60,
... |
196d3d05-f356-40af-928a-8101afbed461 | easy | karanganesan | 2026-08-05 10:00:17.097958+00:00 | succeeded | 7af54e899459a26200a71b0c876d5e3d942c2eb7e1c418be459891c847072686 | 22,081 | null | """Parametric looped-transformer family (P1).
One weight-tied transformer block applied k times in latent space. Config
flags cover four P1 families with one file:
looped recall=0 gated=0 tfilm=0 plain weight-tied loop
looped-recall recall=1 re-inject the input embedding each
... | {
"id": "196d3d05-f356-40af-928a-8101afbed461",
"created_at": "2026-08-05 10:00:17.097958+00:00",
"db_md5": "255e2f927dd274a2579557bd8a91931d",
"submitter": "Karan Ganesan",
"github_login": "karanganesan",
"run_id": "ce26c623-f877-4d36-90b7-9ec8a8bf7f93",
"tier": "easy",
"dataset_id": "e1",
"status": ... | {
"score": {
"mean_loss": 8.95827341079712,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0283333333209157
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 9.388903617858887,
"example_count": 100,
... |
196f9db8-87ea-4ce7-875e-8dbcefbab1ea | easy | saketirl | 2026-08-04 04:24:46.229128+00:00 | succeeded | 7bf30e19d90e7f6cadb4ad67d0b2cd51634a4fd2d29a507be6f840a0dd78f8c1 | 3,628 | null | """Basic single-pass Transformer with PyTorch AdamW."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
assert_model_state,
)
D_MODEL = 128
NUM_HEADS = 4
... | {
"id": "196f9db8-87ea-4ce7-875e-8dbcefbab1ea",
"created_at": "2026-08-04 04:24:46.229128+00:00",
"db_md5": "eaac69fe345f82973fcf392c504a40ff",
"submitter": "saketirl",
"github_login": "saketirl",
"run_id": "8d5fcd48-0a74-44fe-853c-72bee65e8999",
"tier": "easy",
"dataset_id": "e1",
"status": "succeede... | {
"score": {
"mean_loss": 2.971516251564026,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.009999999776482582
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.812049150466919,
"example_count": 100,... |
197287c2-dff4-4b36-b18d-67825fcfd220 | easy | DDanlov | 2026-08-09 00:07:07.183999+00:00 | failed | 2922560cbc5c9dbc9ef51745fc19aeae4d042e9bdaeffefa3d67347001412699 | 15,981 | null | from __future__ import annotations
import sys
import os
import math
import time
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim.optimizer import Optimizer
VOCAB_SIZE = 17
DIGIT_OFFSET = 7
class RohanShampoo(Optimizer):
"""Self-contained RohanShampoo optimizer with eigenvalue ma... | {
"id": "197287c2-dff4-4b36-b18d-67825fcfd220",
"created_at": "2026-08-09 00:07:07.183999+00:00",
"db_md5": "68e0016b327c4d510ffaf73949799038",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "13877a53-ad38-400e-a3b0-2a36396849f4",
"tier": "easy",
"dataset_id": "e5",
"status": "failed",
... | null
|
19736d29-67c7-4555-bff2-b12a23a3d949 | easy | khushidahi | 2026-08-03 19:35:38.550808+00:00 | succeeded | 0a5967d1f4b9b94afcd3acd1d27c9e94d75fbbc044ad7f3a572ae01491507361 | 15,247 | null | """R6 structured decimal-workspace model for One Layer Deeper.
The model does not implement multiplication, modular reduction, or the public
recurrence. It uses the public tokenizer structure to build learned, right-
aligned decimal tapes for N, X, and T, then applies a shared neural transition
to a mutable work tape.... | {
"id": "19736d29-67c7-4555-bff2-b12a23a3d949",
"created_at": "2026-08-03 19:35:38.550808+00:00",
"db_md5": "1bc41e2610eb4fda84fc9ef5c0bd2763",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "53085281-2f52-44aa-a1c6-c056779f037e",
"tier": "easy",
"dataset_id": "e5",
"status": "succ... | {
"score": {
"mean_loss": 2.1975706815719604,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.008333333302289248
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1931111812591553,
"example_count": 60... |
1976d5fe-c8b5-42f6-be00-64b3273da573 | easy | ehonig | 2026-08-19 22:01:06.238815+00:00 | succeeded | 35d7edda988371c95603b4fe3ce83fd818d652fb1c378fa9b575e00d0f83f390 | 19,987 | null | """A plain Transformer for repeated modular squaring, and nothing else yet.
This is a deliberate restart. The previous submission accumulated a field
parser, a carry scan, a learned reciprocal, periodic features, a digit
bottleneck, a depth selector and two cell types, most of them tuned around a
memorisation table th... | {
"id": "1976d5fe-c8b5-42f6-be00-64b3273da573",
"created_at": "2026-08-19 22:01:06.238815+00:00",
"db_md5": "88cc143a5fc4772ab35f18ad71e8e3d3",
"submitter": "Edouardo Honig",
"github_login": "ehonig",
"run_id": "eb6b6ee1-5c02-4d26-9e9d-5ca3e5adddba",
"tier": "easy",
"dataset_id": "e3",
"status": "succ... | {
"score": {
"mean_loss": 2.248311996459961,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.013125000055879354
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.306095838546753,
"example_count": 800,... |
197855e5-96f6-4265-a0c8-9661db042a99 | easy | CodeReclaimers | 2026-08-20 15:19:48.613916+00:00 | succeeded | ecdc5bc3549f21ec4b5703f65802288e92bc05cd2c1587f9346ab4e680aba37d | 42,274 | null | """v39_keyed: keyed-modulus CRT learner — Run A of T1_CERT_PLAN_20260819.
HYPOTHESIS A (alias modulus): h1's displayed N is a stable per-identity
alias phi(M) of a hidden modulus M <= 2^20 from enumerated factor pairs
(<=10-bit primes), while x and y are displayed plainly and the step closes
per prime channel (squarin... | {
"id": "197855e5-96f6-4265-a0c8-9661db042a99",
"created_at": "2026-08-20 15:19:48.613916+00:00",
"db_md5": "4734e439309d100106007d6b654ccab0",
"submitter": "CodeReclaimers",
"github_login": "CodeReclaimers",
"run_id": "d7a2b0e6-5a2a-4a72-99e4-d3b279c23080",
"tier": "easy",
"dataset_id": "e4",
"status... | {
"score": {
"mean_loss": 0.0,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.0,
"example_count": 1200,
"exact_accuracy": 1.0,
... |
197c8206-ed90-4b3f-a038-310032bd0ee2 | easy | G-AshwinKumar | 2026-08-15 21:48:46.222732+00:00 | succeeded | d4979f6129767108399470537bdc1e923cbfae92e66a1ce12983eb5ba63e6136 | 10,493 | null | """Ring Axiomatic TRM (Idea 4): Self-Supervised Ring Homomorphism Regularization.
Enforces algebraic ring consistency on the latent manifold:
1. Idempotency Invariance: Ensures the modular reduction operator is an exact projection.
2. Homomorphic Multiplication: Ensures Latent(a * b mod N) == Latent(a) * Latent(b) mod... | {
"id": "197c8206-ed90-4b3f-a038-310032bd0ee2",
"created_at": "2026-08-15 21:48:46.222732+00:00",
"db_md5": "767c62156504d45f55973705ea60222a",
"submitter": "Ashwin Kumar",
"github_login": "G-AshwinKumar",
"run_id": "d38363fe-1e9e-44c8-acea-0c8f9f693ad5",
"tier": "easy",
"dataset_id": "e1",
"status": ... | {
"score": {
"mean_loss": 9.49474573135376,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.03833333309739828
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 9.790541648864746,
"example_count": 100,
... |
1982aad9-b45c-4467-9488-e7b76c3b9be8 | easy | velocizapkar | 2026-08-06 22:01:22.505968+00:00 | succeeded | 2cdc8df299f838ed46b4dbd180a0501a8b3eb4d7a622f74dc05597fe30ca22ff | 53,088 | null | """Fast packed residual-free Attention-Residual transformer with Muon.
The model is a generic learned sequence model. It contains no parser,
arithmetic primitive, task constant, generated data, label feedback, or answer
construction. Every macro-state is produced by a serial pack of attention and
MLP transformations... | {
"id": "1982aad9-b45c-4467-9488-e7b76c3b9be8",
"created_at": "2026-08-06 22:01:22.505968+00:00",
"db_md5": "9a2489cd6b2a3bd68e11305cc5725b39",
"submitter": "Aakanksh Zarapkar",
"github_login": "velocizapkar",
"run_id": "4941e450-41d6-4614-8988-5b8d878cb8b7",
"tier": "easy",
"dataset_id": "e3",
"statu... | {
"score": {
"mean_loss": 2.8279151916503906,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.018124999478459358
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.8550374507904053,
"example_count": 80... |
19869149-9e97-4be7-b77d-047d6f3428cd | easy | shreyash-chonkie | 2026-09-01 03:29:27.767828+00:00 | succeeded | a31e7251ed1e2910090d4a522bf15f374e699b4d7d9202b356aa481ccebe51da | 20,246 | null | """Learned token scanner feeding a random recurrent path reservoir."""
from __future__ import annotations
from itertools import permutations
import torch
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
assert_model_s... | {
"id": "19869149-9e97-4be7-b77d-047d6f3428cd",
"created_at": "2026-09-01 03:29:27.767828+00:00",
"db_md5": "1f1fc0968b24c6f2eccbda4e932aef45",
"submitter": "Shreyash",
"github_login": "shreyash-chonkie",
"run_id": "92653f58-41e1-442e-ae2b-ed354e892923",
"tier": "easy",
"dataset_id": "e6",
"status": "... | {
"score": {
"mean_loss": 0.0,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.0,
"example_count": 60,
"exact_accuracy": 1.0,
"c... |
198794d1-c3bd-4c62-9850-d78c558e0659 | easy | KaustubhKumar05 | 2026-08-25 15:36:31.495498+00:00 | succeeded | 0dd780da0f2643e57765b3d6e0e2a43ad3d3d1878ba500bfc9e4abda1f9875d1 | 17,167 | null | """conv_abacus: final_v1 + the CORRECT Abacus digit embedding.
final_v1 indexed positions from the end of the whole padded sequence, so the
units digit of N and the units digit of x landed on DIFFERENT embeddings --
the exact failure McLeish et al. (arXiv:2405.17399) identify. This uses a
table indexed by significance... | {
"id": "198794d1-c3bd-4c62-9850-d78c558e0659",
"created_at": "2026-08-25 15:36:31.495498+00:00",
"db_md5": "3009ed5dc9d1fec3670f0014eba79c15",
"submitter": "koz",
"github_login": "KaustubhKumar05",
"run_id": "894d9cb2-31ed-4c10-83a7-bc9026da1ffd",
"tier": "easy",
"dataset_id": "e8",
"status": "succee... | {
"score": {
"mean_loss": 7.500194072723389,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.14385027065873146
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 8.947661399841309,
"example_count": 85,
... |
1988192c-1ca9-4d03-9af3-1c274e508797 | easy | DDanlov | 2026-08-21 20:03:05.010311+00:00 | failed | 3c9f69e0ac774350e847a3e90ff5f764194f24a7ee8d3cfdc1e71d71792daa0c | 25,832 | null | from __future__ import annotations
import math
import time
from typing import Any, Optional, Tuple, Dict
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import Optimizer
try:
from benchmark.api import ModelSpec, OptimizerBundle, OptimizerSpec, Submission, assert_model_state
exce... | {
"id": "1988192c-1ca9-4d03-9af3-1c274e508797",
"created_at": "2026-08-21 20:03:05.010311+00:00",
"db_md5": "cc60e2259118a0f06b5a777ccb6f6c43",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "18d008c7-e4c9-4e08-871a-46605fb5fa26",
"tier": "easy",
"dataset_id": "e5",
"status": "failed",
... | null
|
1990b421-918d-4b07-9d88-bdc336f4d3dc | easy | shirvani-jr | 2026-08-28 09:25:34.667733+00:00 | succeeded | e170f10a49d6e796359014818f78177b4673617dd5807a574a8d70606c509bea | 48,965 | null | """Adaptive learned discrete recurrent fabric for One Layer Deeper.
A gradient-trained, operation-free recurrent computational machine. The
model learns, from the competition's endpoint supervision: how to parse the
prompt into a bank of categorical digit registers, a reusable transition
circuit applied serially, a l... | {
"id": "1990b421-918d-4b07-9d88-bdc336f4d3dc",
"created_at": "2026-08-28 09:25:34.667733+00:00",
"db_md5": "9d68900e603880db87f1ddcc250165f2",
"submitter": "Ali",
"github_login": "shirvani-jr",
"run_id": "e40f009c-4083-4367-abd5-4475b2c3cb4e",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 1.8457187414169312,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.07999999821186066
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.794968605041504,
"example_count": 100,... |
1998dd86-321a-477d-9131-f7fcae160c79 | easy | sapient-sapiens | 2026-08-09 14:42:29.538869+00:00 | succeeded | 8af468f70e7a0a8ca7e2c76f93a9c3477b732273876c44d2b3e7fdcb9490007c | 13,700 | null | """Near-500M adaptive recurrence grown from the hosted-working baseline."""
from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import ModelSpec, OptimizerBundle, OptimizerSpec, Submission, assert_model_state
D_MODEL = 6400
NUM_HE... | {
"id": "1998dd86-321a-477d-9131-f7fcae160c79",
"created_at": "2026-08-09 14:42:29.538869+00:00",
"db_md5": "8315b1cadb0a2cb0510f2255efe377f1",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "dfc3a6be-f839-4cb6-9d9b-b6acf16e5647",
"tier": "easy",
"dataset_id": "e1",
"status": "su... | {
"score": {
"mean_loss": 2.402896161960953,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2735799728934443,
"example_count": 100,
"exac... |
1998ebca-8f8b-4da5-8cff-50fde203b498 | easy | velocizapkar | 2026-08-06 21:56:32.408611+00:00 | succeeded | 881b4a455e11dd912a032ea203b764c090dec9e44c4d0d7106eb675af6fa43d6 | 53,087 | null | """Fast packed residual-free Attention-Residual transformer with Muon.
The model is a generic learned sequence model. It contains no parser,
arithmetic primitive, task constant, generated data, label feedback, or answer
construction. Every macro-state is produced by a serial pack of attention and
MLP transformations... | {
"id": "1998ebca-8f8b-4da5-8cff-50fde203b498",
"created_at": "2026-08-06 21:56:32.408611+00:00",
"db_md5": "39fd922fe1deac8cfbc500c2c4a29074",
"submitter": "Aakanksh Zarapkar",
"github_login": "velocizapkar",
"run_id": "300df7e6-798a-400d-8d8e-db5d4d43db1d",
"tier": "easy",
"dataset_id": "e3",
"statu... | {
"score": {
"mean_loss": 2.4941529035568237,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.00874999980442226
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.576981544494629,
"example_count": 800,... |
199b425d-9014-407a-bfa8-82bb5edaa72c | easy | CodeReclaimers | 2026-08-15 05:01:06.120468+00:00 | succeeded | acaae3e694638e70e9600d2f2a9476aaee199d3cdd4ad6d8df1a49005e3dd588 | 57,228 | null | """v31: merge of the prime-agent low-rank universal-CRT design (hosted M4:
100%, Max T=64, OOD-N Max T=64) with the v30 streamed large-N readout.
From the low-rank design: per-prime residue embeddings E (sum m, DIM) with a
shared readout w (DIM, WMAX) — per-prime cost m*DIM, so the bank covers all
primes 2..4095 (~17M... | {
"id": "199b425d-9014-407a-bfa8-82bb5edaa72c",
"created_at": "2026-08-15 05:01:06.120468+00:00",
"db_md5": "fc08ccee7087a70aeefaaa03c48367d2",
"submitter": "CodeReclaimers",
"github_login": "CodeReclaimers",
"run_id": "ed60360a-7de9-45c6-b160-5794710f519e",
"tier": "easy",
"dataset_id": "e8",
"status... | {
"score": {
"mean_loss": 0.0,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.0,
"example_count": 85,
"exact_accuracy": 1.0,
"c... |
199e0ee4-d31c-4ab1-9053-ab1889990411 | easy | DDanlov | 2026-08-29 14:30:47.982820+00:00 | succeeded | 276baf71354af78017d31df70bfbf18a6dd707f74e86ccec4b8b442c4ed6ca13 | 27,030 | null | """
trial_l10_l3_d1024_k64_rel01.py
L10: L=3, D=1024, K_max=64, rel_force=1% (0.01), progress_thresh=0.75, lr=0.004
"""
import math
import time
import contextlib
from dataclasses import dataclass
from typing import Optional, Tuple, Dict, Any, List, Union
import torch
import torch.nn as nn
import torch.nn.functional as... | {
"id": "199e0ee4-d31c-4ab1-9053-ab1889990411",
"created_at": "2026-08-29 14:30:47.982820+00:00",
"db_md5": "647681ec9aca7e7a94441e4db9e8344c",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "ebe8d601-836a-4dff-b4aa-f975f0223bf7",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.315960358774879,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.008333333333333333
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.334452578305129,
"example_count": 600,... |
199f2155-5ad4-4818-9aec-5b9cb78194c8 | easy | Alan-Qin | 2026-08-12 12:37:29.280895+00:00 | succeeded | 98058feb28203fb39254126f63cd094444d911b52c8e4dcb4524fadc9aa72a57 | 12,898 | null | """Cross-attention looped transformer with a value-assembly front-end (v4).
v3 established a healthy loop skeleton: frozen prompt memory, K answer
tokens iterated by a tied self+cross+MLP block, per-row gather at depth T.
v4 attacks the learning-speed bottleneck identified in the v3 metrics
(still memorizing when the... | {
"id": "199f2155-5ad4-4818-9aec-5b9cb78194c8",
"created_at": "2026-08-12 12:37:29.280895+00:00",
"db_md5": "72f6ba7a52dd167bb2f710b1ff54fcac",
"submitter": "Zeyu Qin",
"github_login": "Alan-Qin",
"run_id": "a5343dfe-57ec-42f1-ba33-7f0d25d31d08",
"tier": "easy",
"dataset_id": "e1",
"status": "succeede... | {
"score": {
"mean_loss": 1.8697091341018677,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.018333333544433117
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.8064814805984497,
"example_count": 10... |
19c47103-93db-4db8-a8ae-8840a5044feb | easy | DDanlov | 2026-08-16 22:14:14.215549+00:00 | succeeded | dd086e4917a2c6344ea0112fd24451bd9074c88cf1f09e0e9526517434b10ece | 17,084 | null | """
Production Port-Hamiltonian RoPE Dynamic Deliberation Submission
Configuration: Champion Run 9 with Batch Size = 128:
- Topology: dim=512, stages=8, heads=16 (head_dim=32), max_macro_passes=6
- Dynamics: dt=0.08, dissipation_scale=0.05, beta_init=0.05
- Position Encoding: DynamicRotaryEmbedding (RoPE, scale-free, u... | {
"id": "19c47103-93db-4db8-a8ae-8840a5044feb",
"created_at": "2026-08-16 22:14:14.215549+00:00",
"db_md5": "16f310c69d1294861ca4e4e48c294f5c",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "d2f811a7-6b43-4da2-8340-b78a18943d10",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 3.2749649338294144,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.014999999987582365
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.630475997924805,
"example_count": 100... |
19cc3522-cbb0-487e-b225-f73ed088f8c7 | easy | DDanlov | 2026-08-13 22:41:59.998475+00:00 | failed | bc4a7bc45b10a3c13b98738aff70c8c4dc8398d26bef4179e6f6bd5aa3533f0c | 22,144 | null | import torch
import torch.nn as nn
import torch.nn.functional as F
import math
from dataclasses import dataclass
from typing import Optional
from torch.optim import Optimizer
try:
from client import Submission, assert_model_state, OptimizerBundle
except ImportError:
Submission = None
assert_model_state = N... | {
"id": "19cc3522-cbb0-487e-b225-f73ed088f8c7",
"created_at": "2026-08-13 22:41:59.998475+00:00",
"db_md5": "0b6da4ec56c8a0c3fa0f0516e6b97e39",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "b53bdc29-24ae-45ef-bc67-f6d5b2918b94",
"tier": "easy",
"dataset_id": "e3",
"status": "failed",
... | null
|
19d2851c-c89c-4f12-a52f-412666194bb5 | easy | DDanlov | 2026-08-20 11:47:07.506321+00:00 | succeeded | 91bcd16ed278b4a724edd1cb7a9ab31cb0a218d3d24d99d1ff29b2648b219228 | 20,914 | null | from __future__ import annotations
import math
import time
from typing import Any, Optional, Tuple
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import Optimizer
try:
from benchmark.api import ModelSpec, OptimizerBundle, OptimizerSpec, Submission, assert_model_state
except Imp... | {
"id": "19d2851c-c89c-4f12-a52f-412666194bb5",
"created_at": "2026-08-20 11:47:07.506321+00:00",
"db_md5": "793806eaccd3e2de6da128686a7d7194",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "95105937-bb9c-4998-b629-5bc7d542a4d6",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.293598382982725,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0129166666790843
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.293537055430391,
"example_count": 600,
... |
19d68735-c948-43a4-bfe3-d48bfdb311b3 | easy | erdavis0 | 2026-08-18 05:54:04.210230+00:00 | succeeded | 6e308c4774094cee70cd73413e69d451b152314db47370b35d0bb4b6eec33dc8 | 34,564 | null | """Canonical token-register transition learner, version 43.
This arm combines the reviewed Pro-sized cross-attention role binder with a
trainable bounded depth head. The head reads only learned prompt-token and
semantic features; T remains outside the shared autonomous transition and no
target, fixed token ID, or arit... | {
"id": "19d68735-c948-43a4-bfe3-d48bfdb311b3",
"created_at": "2026-08-18 05:54:04.210230+00:00",
"db_md5": "fdc468e05d37e51f196e706527ce86a3",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "14d559bf-279a-4400-b872-1d4373f96641",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded",... | {
"score": {
"mean_loss": 2.6362982559986032,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0008333333457509677
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.628765514345982,
"example_count": 60... |
19e27057-4a39-4fcb-84d8-59d46be02bce | easy | alirezashirvani-jr | 2026-09-01 00:25:34.722853+00:00 | succeeded | 904ed8f0729c0cf676dfb7eee54dd253a59187286a2410cfbcf15c3c61bc2bd6 | 20,630 | null | from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
assert_model_state,
)
PAD = 0
MODULUS = 2
VALUE = 3
DEPTH = 4
ANSWER = 5
EO... | {
"id": "19e27057-4a39-4fcb-84d8-59d46be02bce",
"created_at": "2026-09-01 00:25:34.722853+00:00",
"db_md5": "fe0ddbe279556a5944aad106e045b8b6",
"submitter": "alirezashirvani-jr",
"github_login": "alirezashirvani-jr",
"run_id": "877e3b67-fbca-4de7-bb80-6a1a7ac50ce5",
"tier": "easy",
"dataset_id": "e5",
... | {
"score": {
"mean_loss": 2.9902860631470434,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0075
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.990910941709852,
"example_count": 600,
"e... |
19fb4091-c1c4-4341-b588-f8f6c0aeec3d | easy | sapient-sapiens | 2026-08-26 10:52:04.607082+00:00 | failed | ccc63816a27928fbd66d8a24152e3ecca07be2164fb4bfe983029b63f648f98f | 45,619 | null | """Matched hierarchy, operator, N-routing, and optimizer research scaffold.
This is a legitimate learned model. It embeds categorical prompt tokens and
structural field positions; it never parses N, X, or T as numeric values. All
recurrence counts are fixed by ``CampaignConfig`` (or randomized independently
of the i... | {
"id": "19fb4091-c1c4-4341-b588-f8f6c0aeec3d",
"created_at": "2026-08-26 10:52:04.607082+00:00",
"db_md5": "446e52559cbe70451f7860a5a14b9c7a",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "bfc0b7e7-1b72-4a6e-b9ec-1f3920b4879c",
"tier": "easy",
"dataset_id": "e5",
"status": "fa... | null
|
19fecf38-0e0b-4cca-b955-fb104913ce78 | easy | sapient-sapiens | 2026-08-15 08:34:13.908578+00:00 | succeeded | 28b8c3ad9f92779ffc1a2d2038d8bfff5d6d8978e20f4b7d2e607f79a8526f9a | 14,711 | null | """Narrow/deep looped Transformer with optional latent workspace tokens.
Research candidate for sampled-modulus fixed-T=2. The encoder builds immutable
prompt memory. A stack of learned operator blocks updates a mutable latent
stream, and the whole stack is reused for each task recurrence. Optional blank
workspace ... | {
"id": "19fecf38-0e0b-4cca-b955-fb104913ce78",
"created_at": "2026-08-15 08:34:13.908578+00:00",
"db_md5": "2c9e1b46e1498fc84fe972355664d789",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "0c9bcdad-5038-4b26-ba27-8a4d6bfb140e",
"tier": "easy",
"dataset_id": "e4",
"status": "su... | {
"score": {
"mean_loss": 2.064469698903924,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0040625
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.0680124282124153,
"example_count": 1200,
... |
1a02f03d-8791-4778-8d4a-6f8a3b2523aa | easy | Avyukth | 2026-08-11 19:57:52.322135+00:00 | succeeded | fdc6f48bdd101f5cce1a40229b2f51791a7df15993e8d44fe94f8bfc68f60708 | 13,386 | null | """T-adaptive recurrent Transformer: loop count scales with T read from the prompt.
The benchmark family composes a transition function T times. This model bakes
that structure in: it decodes the integer T from the prompt's digit tokens and
applies one shared block BASE + K*T times. Phase embeddings are cyclic (period... | {
"id": "1a02f03d-8791-4778-8d4a-6f8a3b2523aa",
"created_at": "2026-08-11 19:57:52.322135+00:00",
"db_md5": "8fd6e25d1db0146be7c91e426272077e",
"submitter": "Subhrajit Makur",
"github_login": "Avyukth",
"run_id": "d556977f-5eb7-4721-9b78-bb7be448cee2",
"tier": "easy",
"dataset_id": "e1",
"status": "su... | {
"score": {
"mean_loss": 9.59187936782837,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.038333332166075706
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 9.837788581848145,
"example_count": 100,
... |
1a1680cb-7b96-4689-97bb-b83839c46bca | easy | khushidahi | 2026-08-05 20:30:25.643024+00:00 | succeeded | 8a7074590a04c1b14cdbac82126f2ecfd816c839f5bfadbbd3db4334a0e0445b | 16,049 | null | """R6 structured decimal-workspace model for One Layer Deeper.
The model does not implement multiplication, modular reduction, or the public
recurrence. It uses the public tokenizer structure to build learned, right-
aligned decimal tapes for N, X, and T, then applies a shared neural transition
to a mutable work tape.... | {
"id": "1a1680cb-7b96-4689-97bb-b83839c46bca",
"created_at": "2026-08-05 20:30:25.643024+00:00",
"db_md5": "8ba1447dcbc5138c8b1de230d56bdc99",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "e51705b1-53e4-43d1-bae5-ad307800e1b7",
"tier": "easy",
"dataset_id": "e5",
"status": "succ... | {
"score": {
"mean_loss": 2.2136688232421875,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.008333333767950535
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.20992112159729,
"example_count": 600,... |
1a1fffd4-d426-4ba4-9f22-858a1024a372 | easy | viridale | 2026-08-16 03:54:21.988896+00:00 | succeeded | 19d07d9030dcba5927ae45e13f087d19935191a93e686c7680777deaeed611fd | 18,061 | null | """Integer-polynomial boundary quotient with clip-invariant acquisition.
hypothesis: quotienting the T1 coordinate fiber to one autonomous polynomial neuron
and projecting its learned coefficients to integers makes the acquired boundary
compose exactly and transfer across modulus identities.
axis: arch
target:... | {
"id": "1a1fffd4-d426-4ba4-9f22-858a1024a372",
"created_at": "2026-08-16 03:54:21.988896+00:00",
"db_md5": "aa64c42e392b1cd8e58ee153383a4ab4",
"submitter": "priormancer",
"github_login": "viridale",
"run_id": "31868b1c-1a9d-41c9-b924-796f40ea388a",
"tier": "easy",
"dataset_id": "e6",
"status": "succe... | {
"score": {
"mean_loss": 0.15029457211494446,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.15949775278568268,
"example_count": 60,
"ex... |
1a20f4f5-fd5c-46cb-bbaf-ddab5e23c77a | easy | ArkinDharawat | 2026-08-13 17:25:49.903812+00:00 | succeeded | d87efcef95ddba08e527f3f53009598142025f99a5826a7e4c8ad042b9435741 | 7,634 | null | """Universal-Transformer-style looped block, v2: step embed + pos offset + LR schedule + 6 loops.
Extends ``looped_ut_padded`` with four ideas drawn from
Graves 2016 (ACT), Dehghani et al. 2018 (Universal Transformer), and
Merrill & Sabharwal 2025 (log-depth transformers):
1. **Learned per-step (timestep) embedding**... | {
"id": "1a20f4f5-fd5c-46cb-bbaf-ddab5e23c77a",
"created_at": "2026-08-13 17:25:49.903812+00:00",
"db_md5": "c4a6b07700d7bedd1598569ce1b9d5a0",
"submitter": "Arkin Dharawat",
"github_login": "ArkinDharawat",
"run_id": "bc5a9083-6723-4a87-800a-c3314fb4807a",
"tier": "easy",
"dataset_id": "e3",
"status"... | {
"score": {
"mean_loss": 2.246999499735068,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.010000000149011612
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2666880296066023,
"example_count": 800... |
1a210f8c-9f55-4c61-93f3-e33efddca2ce | easy | jordanrubin | 2026-08-03 21:20:27.968165+00:00 | succeeded | 09586f90ffa296e3bdbb25e6a0b059c7579344900d2f7d9eb12b0e902c4848f6 | 15,390 | null | """Bidirectional transformer, D=256 x 3 blocks, grokking-tuned AdamW.
Design notes:
- Inputs are padded to config.max_seq_len inside forward and logits sliced
back, so every training batch presents one static shape (compile/cudagraph
friendly, and uniform kernel shapes even in eager).
- Attention mask gets an iden... | {
"id": "1a210f8c-9f55-4c61-93f3-e33efddca2ce",
"created_at": "2026-08-03 21:20:27.968165+00:00",
"db_md5": "ef7d67d872a65aa96e81ca53991ab27d",
"submitter": "Jordan Rubin",
"github_login": "jordanrubin",
"run_id": "48eb5d10-2554-4874-bdfc-c65c0526d1e6",
"tier": "easy",
"dataset_id": "e2",
"status": "s... | {
"score": {
"mean_loss": 3.1292049247667677,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.21166666457429528
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.849792056244286,
"example_count": 300,... |
1a229b97-0749-45b1-bba6-65bb404632fe | easy | newjordan | 2026-08-26 15:34:51.220461+00:00 | failed | 1a4834929d062d5371bc68f732bb7d313f3728c478465309d4f60275fa928325 | 4,623 | null | """StepperNet v3 — weight-tied recurrent stepper, per-step supervision.
Legal: end-to-end learned (rule 8), recurrence allowed (rule 3), no eval-time
solver (rule 7/14), GPU-only (rule 9). forward takes ONLY (input_ids,
attention_mask) because the runner never hands target_positions; per-step
supervision lives in trai... | {
"id": "1a229b97-0749-45b1-bba6-65bb404632fe",
"created_at": "2026-08-26 15:34:51.220461+00:00",
"db_md5": "ffb9182a1b5743ed34d0084d52bb43b2",
"submitter": "Frosty40",
"github_login": "newjordan",
"run_id": "5c641204-edb7-4b5a-bd9d-976100781996",
"tier": "easy",
"dataset_id": "e1",
"status": "failed"... | null
|
1a287398-92f8-49ab-b8ca-8f83dd121b44 | easy | viridale | 2026-08-15 20:44:50.305154+00:00 | succeeded | 4dfbeb7e6385d2502afe70686c46897f208740e88702e73f12a11707dfa8440a | 27,468 | null | """Width-equivariant learned boundary plus exact categorical fiber.
hypothesis: v649's reusable decimal boundary substrate becomes more trainable at
depth when a disjoint state/depth-conditioned extension is projected to vanish
identically at T1 and optimized in a separate evaluator-owned backward pass.
axis: ... | {
"id": "1a287398-92f8-49ab-b8ca-8f83dd121b44",
"created_at": "2026-08-15 20:44:50.305154+00:00",
"db_md5": "da7f8428e7091a1e01277e6705bdd574",
"submitter": "priormancer",
"github_login": "viridale",
"run_id": "bc120e0d-9d87-4eec-b43a-8bf8af778d31",
"tier": "easy",
"dataset_id": "e6",
"status": "succe... | {
"score": {
"mean_loss": 5.8309900760650635,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.13378379493951797
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 6.823963165283203,
"example_count": 60,
... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.