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 |
|---|---|---|---|---|---|---|---|---|---|---|
0864b2c8-9fcd-4b9e-a600-e23c3c11d357 | easy | oupadhyay | 2026-08-27 08:22:35.290651+00:00 | succeeded | ea20b5620002c551b1a17058a333c36b708cbf64c67d6ca0c23470f6e3b977e6 | 3,720 | null | """Persistent globally mixed polynomial recurrence over a whole-number latent state."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import nn
from benchmark import OptimizerBundle,Submission,TokenLossBatch,assert_model_state
W,D,R=4,32,256
class C:
def __init__(s,vocab_si... | {
"id": "0864b2c8-9fcd-4b9e-a600-e23c3c11d357",
"created_at": "2026-08-27 08:22:35.290651+00:00",
"db_md5": "fe463104166f81961efb26d83407c104",
"submitter": "Ojasw Upadhyay",
"github_login": "oupadhyay",
"run_id": "fd0cd293-88dc-40f9-9af7-f352fcfbf64f",
"tier": "easy",
"dataset_id": "e5",
"status": "s... | {
"score": {
"mean_loss": 8.302531241312902,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.005000000012417634
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 8.05084734433435,
"example_count": 600,
... |
08683fd6-dc9a-4362-85b8-dbc8399beffd | easy | erdavis0 | 2026-08-27 01:16:57.077348+00:00 | succeeded | 795df3f9652030e991e92998273fd7a4939c775c6b6c6c9395c3028f971d4e8b | 31,304 | null | """Typed-prefix reversible transition B4: oracle-weighted step training.
Public delimiter syntax routes opaque N, X, and T fields into three strict
functional roles. X initializes only an evolving auxiliary tape, N initializes
only an invariant environment lane, and T reaches only a learned monotone
prefix controller... | {
"id": "08683fd6-dc9a-4362-85b8-dbc8399beffd",
"created_at": "2026-08-27 01:16:57.077348+00:00",
"db_md5": "86ddb36d905f79e81813767c4d60fc8f",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "d4fe106b-8c1c-4705-a25c-b3abe832bfa1",
"tier": "easy",
"dataset_id": "e10",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.7704047560691833,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.1362020242959261
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.355862021446228,
"example_count": 125,
... |
08738380-f3dd-49f7-8af6-d494647bf7c1 | easy | erdavis0 | 2026-08-25 02:35:43.404920+00:00 | succeeded | c9f21dbfa5e3f75ce670918d20378f38fca944ca903cc99fda41be09be46d0ae | 26,762 | null | """Reversible augmented-state transducer over aligned learned digit tapes.
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 ... | {
"id": "08738380-f3dd-49f7-8af6-d494647bf7c1",
"created_at": "2026-08-25 02:35:43.404920+00:00",
"db_md5": "73ab5c56d5aea3ccc98fb3f3b89f5955",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "a849c530-7b80-4315-af84-5e5025f68cb5",
"tier": "easy",
"dataset_id": "e3",
"status": "succeeded",... | {
"score": {
"mean_loss": 2.703629945433932,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.01062500006519258
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.6795319526898647,
"example_count": 800,... |
087757a8-85c1-48da-800c-6fdff685bcf0 | easy | ArkinDharawat | 2026-08-13 17:07:48.580224+00:00 | succeeded | 6a5b13af61293fc4646bce87e74c442216cb52b23758711eaa3d33c2011b5519 | 7,630 | 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": "087757a8-85c1-48da-800c-6fdff685bcf0",
"created_at": "2026-08-13 17:07:48.580224+00:00",
"db_md5": "b73971c5e9ce8dfba26051097a24a4ed",
"submitter": "Arkin Dharawat",
"github_login": "ArkinDharawat",
"run_id": "89b4e653-6e57-41f3-9db1-ea71883db791",
"tier": "easy",
"dataset_id": "e3",
"status"... | {
"score": {
"mean_loss": 2.3027555585651305,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.014375000041909515
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.3566558808401745,
"example_count": 80... |
08791f98-d7c3-4c9c-bed1-3d42e9dce01b | easy | viridale | 2026-08-03 08:07:48.641423+00:00 | succeeded | 048495b2939ce6b664f9d1321bde100e61102289613174ef7d2ccc079f8919d8 | 124,920 | null | """Signed four-locus source-edit transducers with boundary proof.
hypothesis: the Hard generator is a small source-level edit to initialization,
loop count, recurrent expression, or output observation rather than a
larger positive polynomial. Search the base program, every one-locus edit,
and every pair of... | {
"id": "08791f98-d7c3-4c9c-bed1-3d42e9dce01b",
"created_at": "2026-08-03 08:07:48.641423+00:00",
"db_md5": "2922370d159fa840b7116af586880052",
"submitter": "priormancer",
"github_login": "viridale",
"run_id": "bf5d344d-177e-4f5b-9bef-832136ac16d3",
"tier": "easy",
"dataset_id": "e4",
"status": "succe... | {
"score": {
"mean_loss": 0.0030145460624845253,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.0030145460699592575,
"example_count": 1200,
... |
087a67af-f082-4998-8ecf-c112b45d95e5 | easy | himalalps | 2026-08-07 03:22:36.387131+00:00 | succeeded | cfc21c54a7bb0d2ebf649f6e65716ba11d6a013a30487ecc3835be8c626a103d | 67,596 | null | """Algorithmically constrained latent recurrence for One Layer Deeper.
One masked cross-attention pass creates two independent representations:
``state_0`` from X only and an immutable context from N only. T is excluded
from both representations and is used solely to choose how many times the
learned recurrent MLP tr... | {
"id": "087a67af-f082-4998-8ecf-c112b45d95e5",
"created_at": "2026-08-07 03:22:36.387131+00:00",
"db_md5": "e8aea8545ce7e844e3c127955680e8a0",
"submitter": "Haoyu Tang",
"github_login": "himalalps",
"run_id": "638f42a1-037e-4e40-9429-fc445cb466ad",
"tier": "easy",
"dataset_id": "e5",
"status": "succe... | {
"score": {
"mean_loss": 3.4261883383801393,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.008749999888241291
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.6942453384399414,
"example_count": 60... |
088179c9-497a-44e0-a03a-6163705ccea2 | easy | erdavis0 | 2026-08-23 09:29:32.703573+00:00 | succeeded | 542b2178819561fad0e11cb9fa37a4f38cfc836f4d006099f49ecd24cdd925cd | 20,143 | null | """Sequential single-head learned Mealy sweeper over aligned digit tapes.
Public delimiters only route opaque tokens into separate right-aligned N, X,
and T tapes. One learned cell visits exactly one physical slot at a time,
first right-to-left and then left-to-right. It carries separate interaction
and modulus regi... | {
"id": "088179c9-497a-44e0-a03a-6163705ccea2",
"created_at": "2026-08-23 09:29:32.703573+00:00",
"db_md5": "a5752125d92dc9c5aed85c8a3dad12b5",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "8040cadd-bf53-4a46-9faa-974e74f8491e",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded",... | {
"score": {
"mean_loss": 2.142018699409834,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004166666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.137910345744659,
"example_count": 600,... |
088511bc-f8b1-4348-9067-f995ad4c1e5e | easy | adambrzosko | 2026-08-30 16:05:09.684834+00:00 | succeeded | 8099174425643e4a977abc24404e6913e9ec97ed419106a7308f5c05d5a04c2c | 12,497 | null | """One Layer Deeper — digit-slot recurrent machine, v2.
Prompt: [N] digits(N) [X] digits(x) [T] digits(T); label: the decimal expansion
of x^(2^T) mod N, tail-aligned, so prompt position len-1-s carries the digit of
significance s.
The model parses the three fields onto a grid of slots indexed by (role,
significance)... | {
"id": "088511bc-f8b1-4348-9067-f995ad4c1e5e",
"created_at": "2026-08-30 16:05:09.684834+00:00",
"db_md5": "5e17379179dfc87bf1e3a4253be32791",
"submitter": "Adam Brzosko",
"github_login": "adambrzosko",
"run_id": "e1a58469-36a2-43d6-9fa5-edd639202cbe",
"tier": "easy",
"dataset_id": "e4",
"status": "s... | {
"score": {
"mean_loss": 2.0838485465154077,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0070833333457509685
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.086911478325331,
"example_count": 12... |
0887659b-acee-47f7-98c8-caf2e50200da | easy | oupadhyay | 2026-08-24 00:12:51.128115+00:00 | succeeded | 447ac6fe872d93c1c8a19edcbb6a2827f808e3489b13c306e2555ff1469b7667 | 4,318 | null | """Conditional structured-loss screen: hard50."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import nn
from benchmark import ModelSpec,OptimizerSpec,OptimizerBundle,Submission,TokenLossBatch,assert_model_state
W,D,H=4,256,128
class C:
def __init__(s,vocab_size,ma... | {
"id": "0887659b-acee-47f7-98c8-caf2e50200da",
"created_at": "2026-08-24 00:12:51.128115+00:00",
"db_md5": "e8dc9f5d06ba61a02ea5975dfd9edaae",
"submitter": "Ojasw Upadhyay",
"github_login": "oupadhyay",
"run_id": "1e11a68b-181c-43cd-8384-7b1d40aa5e8a",
"tier": "easy",
"dataset_id": "e5",
"status": "s... | {
"score": {
"mean_loss": 3.1659520860512025,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.007083333358168602
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.7664947039343315,
"example_count": 60... |
0890256e-7c4b-488b-a884-acc6e3bb7a51 | easy | alirezashirvani-jr | 2026-08-30 21:53:07.402503+00:00 | succeeded | 0cc4c85108a191064b58154e00eaad373ff3456fb0e99235094e9abb72186c47 | 24,092 | 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
EXPERT_COUNT = 12
STATE_COUNT = 64
PAIR_COUNT = EXPERT_COUNT * (EXPERT... | {
"id": "0890256e-7c4b-488b-a884-acc6e3bb7a51",
"created_at": "2026-08-30 21:53:07.402503+00:00",
"db_md5": "d80aafc0c2a47d0ff2977b43c9603536",
"submitter": "alirezashirvani-jr",
"github_login": "alirezashirvani-jr",
"run_id": "55aa3657-d128-443d-bb31-a03c861b6745",
"tier": "easy",
"dataset_id": "e3",
... | {
"score": {
"mean_loss": 5.0371763706207275,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.006874999729916453
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.9775004386901855,
"example_count": 80... |
089b031a-5e6d-421e-9ad0-48bc8051ce0f | easy | 0Chris5R | 2026-08-07 09:37:18.797522+00:00 | succeeded | f8ecd1700a956da2158a1772e5fb24cf859a13a13797f2824ad3638974ed79f2 | 30,824 | null | """Shared-digit Neural GPU with late full-objective EMA evaluation.
The model aligns decimal digits by significance, forms a learned low-rank
second-order feature map of the current state, and evolves a two-dimensional
workspace with the original convolutional-GRU equations. Each arithmetic
transition applies the sam... | {
"id": "089b031a-5e6d-421e-9ad0-48bc8051ce0f",
"created_at": "2026-08-07 09:37:18.797522+00:00",
"db_md5": "ee24f5ac33eff1551a0c4120c674cf93",
"submitter": "Chris ",
"github_login": "0Chris5R",
"run_id": "75ea79f2-b4cf-4ea0-9aa5-d8e35a677bcc",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.966683608489693,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.011250000000000001
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.3253761225217127,
"example_count": 600... |
089b716b-9eca-484b-9d1c-1606539d2305 | easy | DDanlov | 2026-08-20 22:58:53.207417+00:00 | failed | 55685715c4d1f18ff6c003d31c9c85037f58bac68498eeed2af9fccbf39e2129 | 21,253 | 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": "089b716b-9eca-484b-9d1c-1606539d2305",
"created_at": "2026-08-20 22:58:53.207417+00:00",
"db_md5": "160baa3e28b1a2142fcd1c254a81d9d9",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "21f86779-17c8-4b51-bc7e-75c436d9c78f",
"tier": "easy",
"dataset_id": "e5",
"status": "failed",
... | null
|
089cc6bd-4297-461c-aedf-f0e0e2b3683a | easy | ddoan | 2026-08-14 12:47:40.880967+00:00 | succeeded | d973f2f3b907b89b4392b159fe74d42794a51253786999740cac5b6f181c83cd | 13,769 | null | """Pinned place codec with bilinear routing and a learned carry scan."""
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_sta... | {
"id": "089cc6bd-4297-461c-aedf-f0e0e2b3683a",
"created_at": "2026-08-14 12:47:40.880967+00:00",
"db_md5": "181f2fede48bdf303afc3a2157a08afc",
"submitter": "Doug Doan",
"github_login": "ddoan",
"run_id": "a3a1c0b0-0626-4341-88cb-71e42c2de357",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 5.410569190979004,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.009999999776482582
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 5.673981189727783,
"example_count": 100,... |
08a60bdf-69f8-423b-8796-31f4e867efb4 | easy | sapient-sapiens | 2026-08-09 14:37:56.912427+00:00 | succeeded | 382275856d281ae070ad581e7f108a466e47498c941a9e0d381383c62c4ed2c7 | 12,956 | 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": "08a60bdf-69f8-423b-8796-31f4e867efb4",
"created_at": "2026-08-09 14:37:56.912427+00:00",
"db_md5": "3569dd21c17477c92d16a8c1ea063b06",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "5b31c696-4d24-4f0b-ae68-d9ad7fb709f4",
"tier": "easy",
"dataset_id": "e1",
"status": "su... | {
"score": {
"mean_loss": 2.2748756173194047,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.08333333333333334
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.220638297164618,
"example_count": 100,... |
08b2cd28-6d53-4948-84d7-6253e02e89dc | easy | ehonig | 2026-08-20 04:48:54.727929+00:00 | succeeded | e44bdbc1b993b1ded8fa85b7638826c4fd03ccb4011ece867d6146c9ed97ab92 | 23,007 | 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": "08b2cd28-6d53-4948-84d7-6253e02e89dc",
"created_at": "2026-08-20 04:48:54.727929+00:00",
"db_md5": "29a1b721133a7a85121e6add806ddf56",
"submitter": "Edouardo Honig",
"github_login": "ehonig",
"run_id": "0e2c32f3-184b-4fd4-9554-263f5263c836",
"tier": "easy",
"dataset_id": "e1",
"status": "succ... | {
"score": {
"mean_loss": 2.18602192401886,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0833333320915699
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.201610565185547,
"example_count": 100,
... |
08b49eb2-ca2a-43de-9083-96904217ea09 | easy | viridale | 2026-08-07 02:57:37.021414+00:00 | succeeded | abc10f9114409144d09417fcfccbff54b97f6339194b584be72544a22b853d8f | 11,201 | null | """Field-agnostic whole-number FoNE with an ordinary learned decoder.
hypothesis: whole-number FoNE ingress is useful but a two-dimensional circular digit
head can bottleneck final-string fitting; an unconstrained learned vocabulary
decoder isolates ingress from the organizer-approved representation.
axis: arc... | {
"id": "08b49eb2-ca2a-43de-9083-96904217ea09",
"created_at": "2026-08-07 02:57:37.021414+00:00",
"db_md5": "05489691cb6a9a046d3b3399e6369b98",
"submitter": "priormancer",
"github_login": "viridale",
"run_id": "9645acbe-3966-4030-bde2-8c1f13805837",
"tier": "easy",
"dataset_id": "e1",
"status": "succe... | {
"score": {
"mean_loss": 2.179598569869995,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0833333320915699
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1925199031829834,
"example_count": 100,
... |
08b7f3ca-0cc4-4fed-b56d-ec35b97e0c49 | easy | Maxwells-Demons | 2026-08-06 04:55:08.534547+00:00 | succeeded | a7f59abf8af30954ecbd98e0d3db08f70eb6b3ff4e14b617ec6da1907d773103 | 13,507 | null | """Value-bottleneck recurrent model, e2 configuration (v13 champion).
Same architecture as the e1 submission (pinned digit head, leading-zero
supervision) plus SAM-style extra parameter updates inside the permitted
between_backward_passes callback (transforms parameters/optimizer state in
place; never calls optimizer.... | {
"id": "08b7f3ca-0cc4-4fed-b56d-ec35b97e0c49",
"created_at": "2026-08-06 04:55:08.534547+00:00",
"db_md5": "851dbaac8ef5bdd154290c47a2f5f3b4",
"submitter": "Maxwells_Ayakashi",
"github_login": "Maxwells-Demons",
"run_id": "caf95270-0e65-44ca-991a-21e529e289a6",
"tier": "easy",
"dataset_id": "e2",
"st... | {
"score": {
"mean_loss": 1.908640742301941,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.6310416758060455
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.8263908624649048,
"example_count": 300,
... |
08c3a70b-94b3-456c-8f41-111db910ec3e | easy | chad-atexpedient | 2026-08-09 20:17:56.978697+00:00 | succeeded | e1ab32a75ab19a26388b883f626cea338f072bcb34a8c6e67933c02a06fea93a | 10,028 | null | """BL5B: BL1 post-400 LR 1e-5 continuation."""
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,
)
LANES = 6
IMMUTABLE_... | {
"id": "08c3a70b-94b3-456c-8f41-111db910ec3e",
"created_at": "2026-08-09 20:17:56.978697+00:00",
"db_md5": "8ec0853469c384ba24b951c36203674f",
"submitter": "chad-atexpedient",
"github_login": "chad-atexpedient",
"run_id": "2fc12434-2a5b-4acd-bdbf-22434827a1d6",
"tier": "easy",
"dataset_id": "e5",
"st... | {
"score": {
"mean_loss": 2.847724966578878,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.005416666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.841408383151341,
"example_count": 600,... |
08c4112f-9aaf-49e6-830e-d26f25312609 | easy | TZZheng | 2026-08-06 21:01:57.284641+00:00 | succeeded | 8ed119744976af18c314e728fd72c5ed9655225dec68493b943a85ac099fa0a6 | 7,300 | 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": "08c4112f-9aaf-49e6-830e-d26f25312609",
"created_at": "2026-08-06 21:01:57.284641+00:00",
"db_md5": "136b65e97351ed5257a1af84a0ed649c",
"submitter": "TZZheng",
"github_login": "TZZheng",
"run_id": "09cee4d6-25de-4e34-96a1-81bcaf1cffb4",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.876821219921112,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.06333333067595959
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.827970504760742,
"example_count": 100,
... |
08d75fe8-7d1d-4c8a-bf04-ddce3f8c3194 | easy | DDanlov | 2026-08-21 21:11:27.827989+00:00 | succeeded | 1f331950eb0215c4ac919e3a7b962d025617a71d91a3c3b832ef535034344525 | 27,612 | 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": "08d75fe8-7d1d-4c8a-bf04-ddce3f8c3194",
"created_at": "2026-08-21 21:11:27.827989+00:00",
"db_md5": "9c6b2924e88db541fd04d471bd82074e",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "d3c3fb0f-c4c0-4b02-978b-6c5c9d3ef8e8",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.8144923465618623,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.005
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.8232969072367577,
"example_count": 600,
"e... |
08db5054-6b8d-4947-92a2-d0208dfb92b3 | easy | liam-gb | 2026-08-17 02:26:46.122260+00:00 | succeeded | 82632945ab7e0c87d203dfc4e2f107a9e7f49bf180d7e486a194d56b896646c1 | 14,535 | null | """charblock: a from-scratch body whose per-prime step has a known closed-form
solution, trained in the regime where that solution is documented to be found
from random initialisation.
Five ingredients, declared as one ARCHITECTURE+REGIME bundle:
1. Per-prime two-layer blocks with a QUADRATIC activation. For each ban... | {
"id": "08db5054-6b8d-4947-92a2-d0208dfb92b3",
"created_at": "2026-08-17 02:26:46.122260+00:00",
"db_md5": "a79f5e7cd40035de9d3d3adb1e10b7b5",
"submitter": "liam-gb",
"github_login": "liam-gb",
"run_id": "6651fad3-d4e4-4f10-986c-bd6937d0d109",
"tier": "easy",
"dataset_id": "e8",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.393961191177368,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.4178698994219303
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.6521403789520264,
"example_count": 85,
... |
08dc1ce0-9eb0-459a-b072-714bd171db4f | easy | ehonig | 2026-08-19 17:43:09.037759+00:00 | succeeded | 7e742e0c88b53f31179a90c4c67124c2ace193e8ee946efd8d35b650fef65b23 | 19,543 | 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": "08dc1ce0-9eb0-459a-b072-714bd171db4f",
"created_at": "2026-08-19 17:43:09.037759+00:00",
"db_md5": "4983a21e6453bca76a482c9c89352a41",
"submitter": "Edouardo Honig",
"github_login": "ehonig",
"run_id": "37075d59-f1da-4792-b899-dbc982881b73",
"tier": "easy",
"dataset_id": "e1",
"status": "succ... | {
"score": {
"mean_loss": 1.8527401685714722,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.06500000134110451
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.8033816814422607,
"example_count": 100... |
08dd602d-80fb-41c7-8834-3693025218c7 | easy | adamsoliev | 2026-08-01 23:06:28.394207+00:00 | succeeded | 4e5e218ebc7924a0b4573afddcc33a08452821c3b304702453f04e37d0f71054 | 35,723 | null | """Dual-factor semigroup-block automaton for One Layer Deeper."""
from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
assert_model_state,
)
from torch import Tensor, nn
WIDTH = ... | {
"id": "08dd602d-80fb-41c7-8834-3693025218c7",
"created_at": "2026-08-01 23:06:28.394207+00:00",
"db_md5": "6522259187ade0b39126e1487cbbcf74",
"submitter": "Adam Soliev",
"github_login": "adamsoliev",
"run_id": "1b3cbcdd-96d8-4553-b14d-25ae6d38b077",
"tier": "easy",
"dataset_id": "e2",
"status": "suc... | {
"score": {
"mean_loss": 0.0018129647942259908,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.001812209258787334,
"example_count": 300,
... |
08e12c93-d731-4301-9037-9072b8ded7a1 | easy | khushidahi | 2026-08-16 01:39:08.408167+00:00 | succeeded | 15e50c6ce5b03d88c3cbf1d4357efa867b4598c04d82ebf77203d6dd236518f8 | 23,230 | null | """R87: stable diagonal Neural-GPU transition.
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 transition... | {
"id": "08e12c93-d731-4301-9037-9072b8ded7a1",
"created_at": "2026-08-16 01:39:08.408167+00:00",
"db_md5": "4f78fdda4cb571ae59a5c743c3ac10a4",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "d7313eaa-12ee-4679-a3c8-33a191d9b232",
"tier": "easy",
"dataset_id": "e5",
"status": "succ... | {
"score": {
"mean_loss": 2.6015245171842256,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.011666666679084302
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.6504481245049445,
"example_count": 60... |
08e31133-ffea-404f-a447-c4fea3cf50c7 | easy | shreyash-chonkie | 2026-08-26 13:35:20.264319+00:00 | succeeded | 9379e565d9b06a797dd5ca1911f81475a0b0ed0fec3baa1418edc63871a593d6 | 10,184 | null | """Closed continuous register machine trained through repeated state noise."""
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_stat... | {
"id": "08e31133-ffea-404f-a447-c4fea3cf50c7",
"created_at": "2026-08-26 13:35:20.264319+00:00",
"db_md5": "551caef0882812e68debda5346b8328a",
"submitter": "Shreyash",
"github_login": "shreyash-chonkie",
"run_id": "9fceeb80-2e81-4dce-94ab-c289c4a7bf8d",
"tier": "easy",
"dataset_id": "e1",
"status": "... | {
"score": {
"mean_loss": 2.058289647102356,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.02666666731238365
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.0574822425842285,
"example_count": 100,... |
08e3c877-428e-48e4-85aa-01b6350ee601 | easy | Maxwells-Demons | 2026-08-13 13:35:48.881117+00:00 | succeeded | 14d654a5972ef58e86df9f1c2ab3e9fdb57992559e3495172c36e1ca305c998d | 119,902 | null | """One shared digit operator, its two free neighbours of x, and two learned maps.
The backbone is unchanged from v47/v96: the state IS the digit one-hot, one
application of the operator is
(state digits, N digits) -> next state digit logits
and the straight-through code of those logits is the whole state handed to... | {
"id": "08e3c877-428e-48e4-85aa-01b6350ee601",
"created_at": "2026-08-13 13:35:48.881117+00:00",
"db_md5": "effbb4c3611b32bef2f97e97c48e4b79",
"submitter": "Maxwells_Ayakashi",
"github_login": "Maxwells-Demons",
"run_id": "80a87fde-b679-4ef0-9b19-8be4026ec1d2",
"tier": "easy",
"dataset_id": "e3",
"st... | {
"score": {
"mean_loss": 4.306274959439258,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.8456250083446503
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.148578056629134,
"example_count": 800,
... |
08eb49f9-fea8-414e-92e6-fde2f36de2e0 | easy | pculaf | 2026-08-15 19:15:11.218720+00:00 | succeeded | a0bde36d4aac49e00f9dc958491dd128097236ccd1034db8a6e2d3270d6e2cfe | 10,278 | null | """Muon-optimized recurrent Transformer for the E6 activation/depth sweep."""
from __future__ import annotations
from collections.abc import Iterable
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submissio... | {
"id": "08eb49f9-fea8-414e-92e6-fde2f36de2e0",
"created_at": "2026-08-15 19:15:11.218720+00:00",
"db_md5": "19eb1c832a1a4e0e53dfdd767f6a4410",
"submitter": "Pavle Culafic",
"github_login": "pculaf",
"run_id": "ec146833-7b7d-4d3b-9ddc-690713169b53",
"tier": "easy",
"dataset_id": "e6",
"status": "succe... | {
"score": {
"mean_loss": 1.909517526626587,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.048648651689291
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.889516830444336,
"example_count": 60,
... |
08ed2a28-52fe-4484-ab57-bf27643214ae | easy | chad-atexpedient | 2026-08-16 04:14:54.395110+00:00 | succeeded | 9675f2c55ace5d2049add9872ce75b70d4a8f00f324de4f0aed715d2074a031c | 21,211 | null | """X55: Hybrid continuation-label + state consistency.
Combines X54's continuation-label supervision with X18's state consistency loss.
- Continuation: supervise A at step a+b with B's label z when A.output == B.input.
- Consistency: match T=2 intermediate state (step 1) with T=1 final state when
T=2.output == T=1.i... | {
"id": "08ed2a28-52fe-4484-ab57-bf27643214ae",
"created_at": "2026-08-16 04:14:54.395110+00:00",
"db_md5": "3683282426a83b75e3eab7287153364c",
"submitter": "chad-atexpedient",
"github_login": "chad-atexpedient",
"run_id": "288f7339-71c1-4941-999c-b231055dcbf7",
"tier": "easy",
"dataset_id": "e5",
"st... | {
"score": {
"mean_loss": 2.1828825919637316,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.009166666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.192739330599661,
"example_count": 600... |
08edafa1-baf6-4690-a650-155158aaf867 | easy | DDanlov | 2026-08-21 21:10:14.208595+00:00 | succeeded | d9c5bbb6a9d38253ee92b0668537812f885247c0b5cfbb57fb25effb9fed53a2 | 27,612 | 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": "08edafa1-baf6-4690-a650-155158aaf867",
"created_at": "2026-08-21 21:10:14.208595+00:00",
"db_md5": "ab327f9e3e9d44ef78d5ba176a750cf2",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "17e93289-5066-47d0-9003-b97451520acf",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.8145657601940757,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004583333333333333
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.8233412795537256,
"example_count": 60... |
08f6252b-eb44-4cf2-813e-9ff23b7e33ce | easy | christopherhwood | 2026-08-05 17:02:26.236204+00:00 | succeeded | 4fc02e215f23a8d9cbaf900b5dff00a5440c5ee327e87f373f9006d1546ccec7 | 14,628 | null | """Endpoint-trained length-general squaring with product-aware ripple scans.
This is the benchmark-compatible descendant of the successful sequential-ripple
diagnostic. It keeps a positional digit register and weight-tied recurrent
scans, exposes the full digit-pair outer product so multiplication does not have
to be... | {
"id": "08f6252b-eb44-4cf2-813e-9ff23b7e33ce",
"created_at": "2026-08-05 17:02:26.236204+00:00",
"db_md5": "438fc59d7999612579c0dd425d2f14fb",
"submitter": "Chris Wood",
"github_login": "christopherhwood",
"run_id": "9b8fb80b-5c0e-480f-8425-2089df70a8a7",
"tier": "easy",
"dataset_id": "e5",
"status":... | {
"score": {
"mean_loss": 2.3295116290812494,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.007916666679084302
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1994459110525155,
"example_count": 60... |
08fe579f-942b-400d-a61a-41cfc4b24eaa | easy | KaustubhKumar05 | 2026-08-29 04:42:13.454302+00:00 | succeeded | b2654f8e0d8745057a52bb0a181968c7ede845dd5833915772b99688f6f4244f | 13,636 | null | """op2_t1w -- stock cell_op2 + T=1 loss upweighting (the scored rung).
Forked from cell_op2. ORIGINAL DOC BELOW.
cell_op2 -- ONE shared squaring operator with internal depth, applied T times.
WHY THIS SHAPE (verified on the real data, not assumed):
answer(T=t) == x squared t times mod N 20000 / 20000 e... | {
"id": "08fe579f-942b-400d-a61a-41cfc4b24eaa",
"created_at": "2026-08-29 04:42:13.454302+00:00",
"db_md5": "acf72ee82224dd4e4c78f0697e83e305",
"submitter": "koz",
"github_login": "KaustubhKumar05",
"run_id": "38739da8-83a5-404b-b2b2-c9d876580e9b",
"tier": "easy",
"dataset_id": "e5",
"status": "succee... | {
"score": {
"mean_loss": 2.853837036965907,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.007916666666666666
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.792262303722279,
"example_count": 600,... |
09064c4a-8eb8-4f38-980d-3484f0ad268f | easy | velocizapkar | 2026-08-06 21:44:22.891751+00:00 | succeeded | 3173103b87dbe19a32aa6dec15f145057ba054b964e517196b3529370121f509 | 52,044 | 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": "09064c4a-8eb8-4f38-980d-3484f0ad268f",
"created_at": "2026-08-06 21:44:22.891751+00:00",
"db_md5": "874fd43fd595548225b74f479cc67e79",
"submitter": "Aakanksh Zarapkar",
"github_login": "velocizapkar",
"run_id": "226c8c13-a618-4187-a5be-2b86ced905fa",
"tier": "easy",
"dataset_id": "e3",
"statu... | {
"score": {
"mean_loss": 2.3987642526626587,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.018749999348074198
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.4334716796875,
"example_count": 800,
... |
09170553-7f45-4073-893e-931c67f21c91 | easy | newjordan | 2026-08-07 03:06:10.408528+00:00 | succeeded | 03f125603411816666ec3c8858138f5bdaa17eb09fbaa839ac7c5bf440b6a35a | 46,384 | null | """Helix-Crawler — a recursive-weight-shared "crawler" ALU for modular squaring.
The task is r_{i+1} = r_i^2 mod n, chained T times: a *recurrence* over a single
operator. This build embodies the crawler / Frugendorff lineage (spark:~/sota_crawler)
mapped onto that structure. n never enters as a table row (unseen modu... | {
"id": "09170553-7f45-4073-893e-931c67f21c91",
"created_at": "2026-08-07 03:06:10.408528+00:00",
"db_md5": "33faba9d68358569abf5258e9060795c",
"submitter": "Frosty40",
"github_login": "newjordan",
"run_id": "7505ce03-985a-4b52-8661-b21a5bb66aaa",
"tier": "easy",
"dataset_id": "e1",
"status": "succeed... | {
"score": {
"mean_loss": 6.461731672286987,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.34833332896232605
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 6.758079528808594,
"example_count": 100,
... |
09255516-4874-46c8-be57-fccb472f1b9a | easy | isaac0804 | 2026-08-07 22:31:27.514445+00:00 | succeeded | d85dfac726dc9b4156f660df0001a68777f9f37a4ad14008bafa9339e0fcc4f0 | 14,004 | null | """E3 candidate: optimizer-side gradient-agreement gating across modulus
subgroups, via bounded multi-pass backward (rule 4).
Every mechanism tried on E3 so far (see docs/findings.md) operates on the
model's forward pass or loss shape: routing, bottlenecks, contrastive
regularizers, curriculum. This one operates purel... | {
"id": "09255516-4874-46c8-be57-fccb472f1b9a",
"created_at": "2026-08-07 22:31:27.514445+00:00",
"db_md5": "5b1ccb123271cdbb7280c17f762dbeff",
"submitter": "Isaac Yong",
"github_login": "isaac0804",
"run_id": "9989d4c8-4ace-4afb-ae75-40ef620aafc1",
"tier": "easy",
"dataset_id": "e3",
"status": "succe... | {
"score": {
"mean_loss": 3.6704562100502525,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.008125000004656614
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.933536235053448,
"example_count": 800... |
092a3b9f-1461-4f14-84f8-544361aa34ee | easy | Mantissagithub | 2026-08-19 10:18:56.628931+00:00 | failed | 63cdb0ede554158429fe3ff3c38bc65519ae3dde1a4fce4ab5b7bd001eab309b | 30,722 | null | """attempt_035 + a contentless T field (depth enters only as iteration count).
ADDENDUM (attempt_042_gated): attempt_040_residual with the cross-iteration
carry scaled by a single learned scalar gate, sigmoid(carry_logit), init 0.5.
This is one scalar parameter (+1 element against the model-state ceiling) and
one mult... | {
"id": "092a3b9f-1461-4f14-84f8-544361aa34ee",
"created_at": "2026-08-19 10:18:56.628931+00:00",
"db_md5": "fd2c2f4599939632103fd9f49f8b6bf2",
"submitter": "Pradheep P",
"github_login": "Mantissagithub",
"run_id": "055f4a6f-49b7-4341-bf56-28f5a73e80d8",
"tier": "easy",
"dataset_id": "e2",
"status": "... | null
|
092c776d-f126-40d8-bdc0-9967dea984da | easy | karanganesan | 2026-08-08 04:14:18.399786+00:00 | succeeded | 8d9eecdc8607c91b190336e24b49597738decc69ce7d4e24ee5930ae97cfcad7 | 27,127 | 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": "092c776d-f126-40d8-bdc0-9967dea984da",
"created_at": "2026-08-08 04:14:18.399786+00:00",
"db_md5": "3e8439ba8ab561733e7dc883fab4ecb2",
"submitter": "Karan Ganesan",
"github_login": "karanganesan",
"run_id": "57ff2ee7-3b27-4021-aed8-74ee9d962320",
"tier": "easy",
"dataset_id": "e3",
"status": ... | {
"score": {
"mean_loss": 8.267003059387207,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004999999771825969
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 9.196698188781738,
"example_count": 800,... |
0932a81d-53d4-4779-86f1-1f1165b84ea6 | easy | viridale | 2026-08-21 20:00:33.489244+00:00 | succeeded | ad1f5a750d3d75998f407ecfac5fe7129bab859b734851aebd564fafe9e27020 | 73,647 | null | """NV045: generic learned per-step position transducer.
hypothesis: NV044 can learn a recurrence, an input/output wrapper, a per-step
digit substitution, and a second-order initial condition, but it cannot move
digit positions *inside* the recurrent state after every step. Add a single
freely learned leng... | {
"id": "0932a81d-53d4-4779-86f1-1f1165b84ea6",
"created_at": "2026-08-21 20:00:33.489244+00:00",
"db_md5": "5641d417776dae9fe4ac0fd49f666f59",
"submitter": "priormancer",
"github_login": "viridale",
"run_id": "91cc20b2-3a90-4a31-8f03-a27db8277d76",
"tier": "easy",
"dataset_id": "e1",
"status": "succe... | {
"score": {
"mean_loss": 2.384185567930013e-07,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.3841855067985307e-07,
"example_count": 100,
... |
09558081-7660-46b1-b10d-48a9b73fd82a | easy | ddoan | 2026-08-14 13:16:49.186326+00:00 | succeeded | ca76428e5347a5af2a33fe60019308cb4bc72ad3a5a16310ba34d2eadc0c2c61 | 13,366 | null | """Pinned place codec with bilinear routing and a learned carry scan."""
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,
TokenLossBatc... | {
"id": "09558081-7660-46b1-b10d-48a9b73fd82a",
"created_at": "2026-08-14 13:16:49.186326+00:00",
"db_md5": "fba786fa241b9ba1f1294789dc85b013",
"submitter": "Doug Doan",
"github_login": "ddoan",
"run_id": "42f588ac-06c7-4550-91e1-43207c50d99f",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.8738192319869995,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.026666666381061077
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.668853282928467,
"example_count": 100... |
095e93fe-d2d1-4370-8588-981d87103243 | easy | AFeinfield | 2026-08-28 00:41:40.467502+00:00 | succeeded | 038a5caa1271f887cd410bc7ef8a781fa6d88a6df1a56a663f0a92c42ff04759 | 15,150 | null | """176: 174 + MINIMAL LOCAL-N CONTEXT (one scalar feature).
175 showed full context lanes break the loss floor by ciphering (train
up, test down). The minimal arithmetic-relevant context is N's own digit
at each place, as one exact scalar into the carry op — necessary for
reduction-aware carries, useless for row ident... | {
"id": "095e93fe-d2d1-4370-8588-981d87103243",
"created_at": "2026-08-28 00:41:40.467502+00:00",
"db_md5": "d7d806490d3d93e4d6391afa77f9c8e0",
"submitter": "AFeinfield",
"github_login": "AFeinfield",
"run_id": "a390053a-e74d-47f4-a960-0858b7e21ef2",
"tier": "easy",
"dataset_id": "e5",
"status": "succ... | {
"score": {
"mean_loss": 2.272403412607005,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.012083333345750968
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2726941707422914,
"example_count": 600... |
0962c8ab-0acb-4973-9a49-432d986f45b1 | easy | chad-atexpedient | 2026-08-16 04:13:00.991765+00:00 | succeeded | bbe8dd4e772b484089a9131e4727bdb75aee3dccc34c37bc7d031224b3709415 | 21,211 | null | """X55: Hybrid continuation-label + state consistency.
Combines X54's continuation-label supervision with X18's state consistency loss.
- Continuation: supervise A at step a+b with B's label z when A.output == B.input.
- Consistency: match T=2 intermediate state (step 1) with T=1 final state when
T=2.output == T=1.i... | {
"id": "0962c8ab-0acb-4973-9a49-432d986f45b1",
"created_at": "2026-08-16 04:13:00.991765+00:00",
"db_md5": "fea1b7a97135f8601a3c4f93e63f4f01",
"submitter": "chad-atexpedient",
"github_login": "chad-atexpedient",
"run_id": "60d43b2e-82fe-4972-8cc8-d597a2f1c4e2",
"tier": "easy",
"dataset_id": "e5",
"st... | {
"score": {
"mean_loss": 2.1785383739032773,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.013333333333333332
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1863481528021294,
"example_count": 60... |
09664a60-6000-4b0c-ae0e-39c455518618 | easy | jamaliki | 2026-08-07 15:33:20.804460+00:00 | succeeded | d66339f74308a5484f0f50941895bc9b7a5ba9d75d1b6ca45bbecb7c82e107e3 | 97,739 | null | """One Layer Deeper submission: clocked latent loop with FSQ restoration.
Design: an encoder reads the input once into a memory. A weight-tied core is
applied m times per macro-step over [register | scratch] slots, cross-attending
to the memory (input re-injection). At each clock edge, registers are snapped
through fi... | {
"id": "09664a60-6000-4b0c-ae0e-39c455518618",
"created_at": "2026-08-07 15:33:20.804460+00:00",
"db_md5": "29d54a05760d5a8d882471169094cdf9",
"submitter": "Kiarash Jamali",
"github_login": "jamaliki",
"run_id": "45c2c6da-782a-407e-a969-7eda23571975",
"tier": "easy",
"dataset_id": "e1",
"status": "su... | {
"score": {
"mean_loss": 3.7551183700561523,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.35499999672174454
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.232954978942871,
"example_count": 100,... |
0966ff8e-57d2-4e89-ab83-2beb8649939b | easy | AFeinfield | 2026-08-27 23:16:58.335691+00:00 | succeeded | bed45df6a5e74ba1a96ec0b860506c21a54c173d053aa21c4ad9b113c581d339 | 13,857 | null | """163: 161 + lr 6e-3 retest
159/160 failed hosted with NaN: bounded features removed the loss-side
pressure valve on place magnitudes, so iterated squaring inflated the
state to inf and sin(inf)=NaN. Clamping the state after carry
normalization keeps every intermediate finite regardless of where
training is in operat... | {
"id": "0966ff8e-57d2-4e89-ab83-2beb8649939b",
"created_at": "2026-08-27 23:16:58.335691+00:00",
"db_md5": "0d9ce9b03d66e1b2034205fa9a4632d3",
"submitter": "AFeinfield",
"github_login": "AFeinfield",
"run_id": "d3b75472-2d76-4537-8ed0-4ea56bb112f3",
"tier": "easy",
"dataset_id": "e1",
"status": "succ... | {
"score": {
"mean_loss": 2.1112091541290283,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0833333320915699
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.11879301071167,
"example_count": 100,
... |
096d8e47-7ce8-4dbe-aacd-411333f3fd28 | easy | alirezashirvani-jr | 2026-08-30 00:20:19.149717+00:00 | succeeded | 672293a76323da62e296a3ee3f10c6ca38ac913874438ad7921545a2384944f0 | 31,682 | 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": "096d8e47-7ce8-4dbe-aacd-411333f3fd28",
"created_at": "2026-08-30 00:20:19.149717+00:00",
"db_md5": "310e8a85d3554c9bd27bdd0854d43589",
"submitter": "alirezashirvani-jr",
"github_login": "alirezashirvani-jr",
"run_id": "b2255799-85ae-4157-88f3-ca832bcf6e63",
"tier": "easy",
"dataset_id": "e9",
... | {
"score": {
"mean_loss": 1.4992516040802002,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.647777795791626
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.1052006483078003,
"example_count": 90,
... |
097bb331-3a26-4fc6-8526-e6e3e6bbf256 | easy | erdavis0 | 2026-08-22 11:43:35.363209+00:00 | succeeded | 50ebc0bceeecde469d62f859deeeb974ac8fb9a872242389d9c0c31e5cafa1df | 8,147 | null | """Ordered scan fields with multiplicative prompt-role fusion."""
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 = 72
FIELDS =... | {
"id": "097bb331-3a26-4fc6-8526-e6e3e6bbf256",
"created_at": "2026-08-22 11:43:35.363209+00:00",
"db_md5": "2db3a330616b6c90d0ff3cea3503178e",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "0cce2490-229a-46c6-881c-2276d38a10e0",
"tier": "easy",
"dataset_id": "e10",
"status": "succeeded"... | {
"score": {
"mean_loss": 8.139312982559204,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.012000000104308128
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 7.271167278289795,
"example_count": 125,... |
0980fb2e-5dab-452b-a0b2-e2268d9f131c | easy | richardcepka | 2026-08-13 23:07:22.085674+00:00 | succeeded | a84a04ac86311cc04fd0977703678602eb06f5dd71cdccae9f866ab97391a1a3 | 16,676 | null | """Depth-recurrent digit model for One Layer Deeper.
The prompt is encoded once, then injected into a shared transformer core on
every recurrent step. A small decimal register is the persistent answer state;
continuous scratch tokens carry everything that is not yet ready to become a
digit. The register stays soft d... | {
"id": "0980fb2e-5dab-452b-a0b2-e2268d9f131c",
"created_at": "2026-08-13 23:07:22.085674+00:00",
"db_md5": "44a9c1b01dea64b345874de8ef157882",
"submitter": "Richard Cepka",
"github_login": "richardcepka",
"run_id": "7171db02-334a-418b-83ed-7348eed34ee1",
"tier": "easy",
"dataset_id": "e5",
"status": ... | {
"score": {
"mean_loss": 2.2191501308803083,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.008749999888241291
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.225773334503174,
"example_count": 600... |
0981128e-9a49-48a1-806b-59cfbaef35e7 | easy | sankalp1999 | 2026-08-04 12:02:48.822305+00:00 | succeeded | 1f4f0bbb3095ddc8fbdcba637637a31caba961ee70dc8509ea256495aef10021 | 8,285 | null | """Bidirectional transformer for the squaring task, looped or plain.
This is the general form: attention over whatever tokens arrive, so a larger
modulus with more digits needs no code change, unlike a flattened-grid MLP
that ties every absolute slot to its own weights.
LOOPED=1 applies ONE weight-shared block T time... | {
"id": "0981128e-9a49-48a1-806b-59cfbaef35e7",
"created_at": "2026-08-04 12:02:48.822305+00:00",
"db_md5": "9832b785079b52ce85f853bfda4d792a",
"submitter": "Sankalp ",
"github_login": "sankalp1999",
"run_id": "6649e720-6e1f-43ca-88f0-4c1e197fdd31",
"tier": "easy",
"dataset_id": "e1",
"status": "succe... | {
"score": {
"mean_loss": 8.17706298828125,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.026666666381061077
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 8.587454795837402,
"example_count": 100,
... |
09869504-cbbc-40b9-82a9-5103591f42ac | easy | alirezashirvani-jr | 2026-09-01 00:37:57.889450+00:00 | succeeded | f0f8d1f2c9edb64aabbd13c10686f8160e7493968812a9010bd8cc6bdcd18472 | 19,002 | 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
EOS = 6
DIGIT0 = 7
DIGIT_COUNT ... | {
"id": "09869504-cbbc-40b9-82a9-5103591f42ac",
"created_at": "2026-09-01 00:37:57.889450+00:00",
"db_md5": "5576a4bf22daaa1273725f80a4751ddd",
"submitter": "alirezashirvani-jr",
"github_login": "alirezashirvani-jr",
"run_id": "10b1fdd7-7e60-4aec-938c-d97295a68bc4",
"tier": "easy",
"dataset_id": "e5",
... | {
"score": {
"mean_loss": 2.851321744070403,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004166666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.879081003067205,
"example_count": 600,... |
098722fb-bb7d-4480-a278-480c65bba038 | easy | sapient-sapiens | 2026-08-19 11:41:59.309135+00:00 | succeeded | f224ab886806559975ff409e9ba9db6263c308bdce59d8da0c8c256bf54f0e20 | 33,256 | null | """Depth-quantized recurrence with a mixture-over-depth objective.
The recurrent state is a bank of right-aligned digit slots. Every operator
application is followed by a soft quantization back onto the token simplex, and
the digit logits produced by that quantization are the answer logits at that
depth. There is no s... | {
"id": "098722fb-bb7d-4480-a278-480c65bba038",
"created_at": "2026-08-19 11:41:59.309135+00:00",
"db_md5": "16b4de7b623067d6a8c0da863125463c",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "8935a83c-3f34-479d-821c-840ab5847f79",
"tier": "easy",
"dataset_id": "e9",
"status": "su... | {
"score": {
"mean_loss": 4.857205867767334,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.014444444794207811
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.814725875854492,
"example_count": 90,
... |
0993e7c4-b04b-46da-9f78-049a4f989673 | easy | shreyash-chonkie | 2026-09-01 03:01:18.132571+00:00 | succeeded | eb0a8cc4da2cdf1236590abe4cf0e5139a9bcbcbd9834f7f8350d2daef24c05c | 21,516 | 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": "0993e7c4-b04b-46da-9f78-049a4f989673",
"created_at": "2026-09-01 03:01:18.132571+00:00",
"db_md5": "f169d321264f929aa2fef59a1b09ea71",
"submitter": "Shreyash",
"github_login": "shreyash-chonkie",
"run_id": "812e673d-869e-4e0f-951b-e5a35d901e4d",
"tier": "easy",
"dataset_id": "e1",
"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": 100,
"exact_accuracy": 1.0,
"... |
0994fba6-3b57-4f17-9336-f9c12e06ed5c | easy | ehonig | 2026-08-14 00:46:40.701684+00:00 | succeeded | 9e5e56489ae8107edcdf83d0d734998ad754f452d69896a853adc6b68b87273d | 48,080 | null | """DigitLoop — a place-value recurrent model for repeated modular squaring.
The task is ``y = x^(2^T) mod N`` with the answer read back tail aligned, one
decimal place value per token position from the right. A stack of attention and
MLP layers has no primitive for the two operations this needs — a partial-product
co... | {
"id": "0994fba6-3b57-4f17-9336-f9c12e06ed5c",
"created_at": "2026-08-14 00:46:40.701684+00:00",
"db_md5": "ae532c7cced822a3232c69468bc3ecf5",
"submitter": "Edouardo Honig",
"github_login": "ehonig",
"run_id": "d430f49a-bd43-4a26-89cd-cb48a405f5ca",
"tier": "easy",
"dataset_id": "e1",
"status": "succ... | {
"score": {
"mean_loss": 5.702927589416504,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.12666666507720947
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 7.400343894958496,
"example_count": 100,
... |
0996a0f4-3225-4691-9053-71ec625dbb72 | easy | lzy54 | 2026-08-29 08:42:56.130432+00:00 | succeeded | f967c3bfea2806a17090394364d66f815812d86d38acc90cb7b0b8126f881c01 | 4,539 | null | """Phase 2 C2: shared block x4 with residual concat 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 = 128
NUM_HEAD... | {
"id": "0996a0f4-3225-4691-9053-71ec625dbb72",
"created_at": "2026-08-29 08:42:56.130432+00:00",
"db_md5": "649789da2b5f7837ff738282292ec92c",
"submitter": "lzy54",
"github_login": "lzy54",
"run_id": "db3863f6-5d8f-4687-977f-346145ab170a",
"tier": "easy",
"dataset_id": "e8",
"status": "succeeded",
... | {
"score": {
"mean_loss": 2.647968053817749,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.04077540151774883
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.5788333415985107,
"example_count": 85,
... |
099f70de-ef4e-4c77-94d6-12d8e5f7589f | easy | 0Chris5R | 2026-08-07 12:29:20.281087+00:00 | succeeded | 571c52ab57f4694312e7020132a3d6982805b5b4ea94cd4abeeaf38c8f6cabe2 | 20,840 | null | """Shared-digit Neural GPU with a compute-matched operator curriculum.
The model aligns decimal digits by significance, forms a learned low-rank
second-order feature map of the current state, and evolves a two-dimensional
workspace with the original convolutional-GRU equations. Each arithmetic
transition applies the ... | {
"id": "099f70de-ef4e-4c77-94d6-12d8e5f7589f",
"created_at": "2026-08-07 12:29:20.281087+00:00",
"db_md5": "c73715ba5893f13987aa1aefc26dcaee",
"submitter": "Chris ",
"github_login": "0Chris5R",
"run_id": "bfedd140-a9be-4d50-ad25-4c12b4506eb8",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 3.1051681513662883,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.017500000024835267
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.1098544266191834,
"example_count": 60... |
09a0bfa3-41a3-4f7a-8a50-5517b3245120 | easy | yashkant | 2026-08-15 13:59:42.293657+00:00 | succeeded | 2a920c54a3b75021f26183100b2c3c8babdcdc24886715ec4e1a11e6be85ad1c | 20,745 | null | """Round1956 Round493 block-two shared-head supervision.
The exact Round493 model, initialization, optimizer, and final prediction path
remain intact. During training both arms additionally read out the hidden state
after encoder block two with the already-existing norm, attention pool,
decoder, and answer head. The c... | {
"id": "09a0bfa3-41a3-4f7a-8a50-5517b3245120",
"created_at": "2026-08-15 13:59:42.293657+00:00",
"db_md5": "b3bf544ca21f3ce76ceb8742d748c33a",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "92c9230c-9558-40c7-a9cd-c115f993b7a5",
"tier": "easy",
"dataset_id": "e2",
"status": "succeed... | {
"score": {
"mean_loss": 5.601670563220978,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.41604167222976685
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 10.158596992492676,
"example_count": 300,... |
09b5416a-8030-408b-9cbe-557e6afc0592 | easy | erdavis0 | 2026-08-19 08:46:37.744516+00:00 | succeeded | 7ef166017dce10b8421c57ae518af1658d74b82dfaf1497b4512d7154140a749 | 9,290 | null | """Raw-vocabulary recurrent operator with identity state decoding."""
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": "09b5416a-8030-408b-9cbe-557e6afc0592",
"created_at": "2026-08-19 08:46:37.744516+00:00",
"db_md5": "e44becb8207f8dd37af235cdc45557ce",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "6f655229-f921-4c10-89db-0827b4c637a7",
"tier": "easy",
"dataset_id": "e8",
"status": "succeeded",... | {
"score": {
"mean_loss": 2.106655478477478,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.05213903822004795
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.0766472816467285,
"example_count": 85,
... |
09b76d1b-0b62-4e98-acf1-0525b15d5d88 | easy | sapient-sapiens | 2026-08-21 21:52:47.516610+00:00 | succeeded | bb6c73eec1db1b1908fcefd87bb21d7d4865263da17ebe2d99965e7faf5019a3 | 33,782 | null | """Depth-quantized recurrence with a mixture-over-depth objective.
The recurrent state is a bank of right-aligned digit slots. Every operator
application is followed by a soft quantization back onto the token simplex, and
the digit logits produced by that quantization are the answer logits at that
depth. There is no s... | {
"id": "09b76d1b-0b62-4e98-acf1-0525b15d5d88",
"created_at": "2026-08-21 21:52:47.516610+00:00",
"db_md5": "2d8bcbd78994b07c34ee529909a34f6c",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "b17a32b5-ffef-48b8-bca1-3343fcc069c9",
"tier": "easy",
"dataset_id": "e5",
"status": "su... | {
"score": {
"mean_loss": 2.818006801794793,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.007083333333333334
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.8095517091686952,
"example_count": 600... |
09b9bb63-a16f-48dc-b6c4-87e1976a8355 | easy | khushidahi | 2026-08-17 15:54:33.599459+00:00 | succeeded | 7c9638217bcb76db95708006269d839b1e757dcb014336ad53a5ee46f1da001e | 23,985 | null | """R205: R201 with a 512-example training batch.
This is the batch-size arm of a preregistered 2x2 factorial. It is identical
to R201's no-recall, width-scaled depth-7 serial Transformer except that the
training batch is 512 instead of 256. The intervention tests sample and H100
throughput without changing the learn... | {
"id": "09b9bb63-a16f-48dc-b6c4-87e1976a8355",
"created_at": "2026-08-17 15:54:33.599459+00:00",
"db_md5": "df092118dd577e8ca59eb125deaab777",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "d33bafa0-305a-4d0d-a8f7-2f837a6e507d",
"tier": "easy",
"dataset_id": "e6",
"status": "succ... | {
"score": {
"mean_loss": 2.4259592294692993,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.021846847608685493
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.4290969371795654,
"example_count": 60... |
09c063de-6009-4fbf-9998-6a82bca95560 | easy | DDanlov | 2026-08-31 20:37:43.471985+00:00 | succeeded | 2bfcc522c89908299c03f5c9c6350edd479aa2ec92c458cca457d035bf3462db | 32,407 | null | """
Unified Continuous Hopfield Equilibrium Propagation Deliberation Engine
=======================================================================
Architecture:
- Deliberation: Continuous Hopfield Multi-Layer Energy System (L=4, D=768, d_ffn=2304, H=16)
- Integration: In-Place Strang-Split Conformal Symplectic Velocit... | {
"id": "09c063de-6009-4fbf-9998-6a82bca95560",
"created_at": "2026-08-31 20:37:43.471985+00:00",
"db_md5": "3eb8d5933b0208081688140ca5ba6351",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "0be9172b-8ec6-4c0e-b498-ecacfd449789",
"tier": "easy",
"dataset_id": "e4",
"status": "succeeded"... | {
"score": {
"mean_loss": 3.2102297621457456,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0011458333333333333
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.2269233803162245,
"example_count": 1... |
09c390b1-dc90-4fa4-84e5-c0385a205159 | easy | ericyuegu | 2026-08-13 19:51:21.291443+00:00 | succeeded | 87939c739ced66540129a0f666266a1ea6a95f69aa4ab8fb7292e1c97aa4b2be | 9,404 | null | """Tiny Recursive Model adapted to the Easy tier's short training budget.
The model keeps separate answer and latent states. Each supervised refinement
runs one H cycle without autograd and one complete H cycle with autograd. The
states are detached between the two refinements, so both predictions supervise
the share... | {
"id": "09c390b1-dc90-4fa4-84e5-c0385a205159",
"created_at": "2026-08-13 19:51:21.291443+00:00",
"db_md5": "b06341e554770e17fdaa6f28f1e81644",
"submitter": "Eric Gu",
"github_login": "ericyuegu",
"run_id": "b3819233-0f53-4723-b562-252df68644f0",
"tier": "easy",
"dataset_id": "e1",
"status": "succeede... | {
"score": {
"mean_loss": 1.997054636478424,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0283333333209157
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.9665967226028442,
"example_count": 100,
... |
09ca6673-d3a1-466d-a11f-323f61423af5 | easy | yashkant | 2026-08-13 00:38:02.697404+00:00 | succeeded | 7c74d210aa672064f4f306c1fd5b05638c34d41cbd7b4018e446da28f4ea6f7d | 40,954 | null | """Round1786 evolutionary crossover of two controlled survivors.
The Round1780 W8/D2 carrier keeps its winning contrastive, exactly
antisymmetric learned set codes. The candidate additionally removes each
initial and recurrent Dykstra correction residual's coordinate mean before
the unchanged tanh bound, combining Ro... | {
"id": "09ca6673-d3a1-466d-a11f-323f61423af5",
"created_at": "2026-08-13 00:38:02.697404+00:00",
"db_md5": "6b2852edb9c63d59879036ff93b71fcd",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "54cf3a7d-4746-402c-a3bf-7e2650b3d068",
"tier": "easy",
"dataset_id": "e3",
"status": "succeed... | {
"score": {
"mean_loss": 2.1237272412341817,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.017500000130385163
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.123758368096935,
"example_count": 800... |
09d02ded-15ec-42e7-be54-95ab412372e1 | easy | k-penchev | 2026-08-23 00:15:02.617756+00:00 | succeeded | cfff0cc5c9c7477d3861e6928726db69121daa8f60240dbb7284122ae5ee1980 | 9,077 | null | """Schedule-Free recurrence with a light worst-digit-aware loss."""
from __future__ import annotations
from collections.abc import Iterable
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
Tok... | {
"id": "09d02ded-15ec-42e7-be54-95ab412372e1",
"created_at": "2026-08-23 00:15:02.617756+00:00",
"db_md5": "b4ecd1aa7ba8f77ab3c034a71cab832f",
"submitter": "Kaloyan Penchev",
"github_login": "k-penchev",
"run_id": "ab5a72d1-44f3-455d-8a4d-891a9b55a8b0",
"tier": "easy",
"dataset_id": "e1",
"status": "... | {
"score": {
"mean_loss": 2.5030837059020996,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.07666666433215141
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.978015422821045,
"example_count": 100,... |
09d83037-bff6-42f7-890b-9d03f37a7571 | easy | pepinu | 2026-08-26 17:44:56.983725+00:00 | failed | 74307ba76d035d71e8330e51db839434ca24ae8d3efbab3f4b12dd2ec11dd273 | 46,270 | null | """Squaring mod N as a learned permutation on CRT components.
WHY THIS SHAPE. (Z_N, x) is a commutative monoid, so every finite-dimensional
linear representation splits into multiplicative characters and squaring is
diagonal there -- the component basis is the only one in which this operation is
shallow. Two consequen... | {
"id": "09d83037-bff6-42f7-890b-9d03f37a7571",
"created_at": "2026-08-26 17:44:56.983725+00:00",
"db_md5": "d7f34595d7f996ce3ab1db374501668a",
"submitter": "pepinu",
"github_login": "pepinu",
"run_id": "f944c15a-50c0-4c80-a8c4-564e0166dcd7",
"tier": "easy",
"dataset_id": "e7",
"status": "failed",
"... | null
|
09df7191-380c-4ff7-b36a-a8efb97c4f8d | easy | isaac0804 | 2026-08-10 21:31:16.421681+00:00 | succeeded | 3a0ab7e3cc2b44b7eed32b02378df701738c124bf6f5c7989b92ff6977a76520 | 8,544 | null | """Clean loop (no damped gate, no step-embedding) + PLAIN static
Hyper-Connections (N=4 parallel residual streams), WITHOUT mHC's
doubly-stochastic constraint -- the un-stabilized base version, submitted
alongside `looped_clean_hyperconnections` (mHC variant) and
`looped_clean_xhc` (xHC variant) to isolate whether the ... | {
"id": "09df7191-380c-4ff7-b36a-a8efb97c4f8d",
"created_at": "2026-08-10 21:31:16.421681+00:00",
"db_md5": "402a78c3e3b0d4da0610e3a313421268",
"submitter": "Isaac Yong",
"github_login": "isaac0804",
"run_id": "8d2f3579-701b-4d27-a39d-8f6f07347ab9",
"tier": "easy",
"dataset_id": "e3",
"status": "succe... | {
"score": {
"mean_loss": 2.3906125496817427,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.013750000153668226
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.406531444055271,
"example_count": 800... |
09f071a1-c3aa-4509-b17f-0f53ad401229 | easy | yashkant | 2026-08-17 00:41:19.211123+00:00 | succeeded | 40184f8861c4649808c05e9ddd3bfafbca29c57f35aff111085ac5da1984191b | 49,530 | null | """Round2217 adds a zero-start square/reduce/answer recurrence.
The confirmed correction-reset model remains intact. Its width-8
answer state is refined by one shared transition with explicit product,
reduction, and answer roles. Immutable categorical N and X summaries
are reinjected on every call and physical calls... | {
"id": "09f071a1-c3aa-4509-b17f-0f53ad401229",
"created_at": "2026-08-17 00:41:19.211123+00:00",
"db_md5": "f17b945fa5b3863a5efdac21e4eba689",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "ac983cf3-3990-4504-b7a3-16d3a0454b1b",
"tier": "easy",
"dataset_id": "e3",
"status": "succeed... | {
"score": {
"mean_loss": 2.132422506843432,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.014375000023283065
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.133055996483597,
"example_count": 800,... |
09f87d64-4546-42c0-85cb-3b8cbf078e56 | easy | erdavis0 | 2026-08-27 01:04:07.987413+00:00 | succeeded | 795df3f9652030e991e92998273fd7a4939c775c6b6c6c9395c3028f971d4e8b | 31,304 | null | """Typed-prefix reversible transition B4: oracle-weighted step training.
Public delimiter syntax routes opaque N, X, and T fields into three strict
functional roles. X initializes only an evolving auxiliary tape, N initializes
only an invariant environment lane, and T reaches only a learned monotone
prefix controller... | {
"id": "09f87d64-4546-42c0-85cb-3b8cbf078e56",
"created_at": "2026-08-27 01:04:07.987413+00:00",
"db_md5": "86ddb36d905f79e81813767c4d60fc8f",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "cb4a8845-787d-4fb3-82d2-9d47bd35a6e2",
"tier": "easy",
"dataset_id": "e8",
"status": "succeeded",... | {
"score": {
"mean_loss": 4.121237874031067,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.13199643790721893
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.427551031112671,
"example_count": 85,
... |
0a016c5a-9549-48ca-ad82-d6df76d47cb4 | easy | sapient-sapiens | 2026-08-14 17:25:32.436454+00:00 | succeeded | 28ba469aae3e42ca679270105f219b6bd47433fa80f83dc154b3c3e55e44c7c3 | 14,299 | null | """Digit-space recurrent operator for repeated modular squaring.
Nothing here converts a token into a decimal number. Digits stay as embeddings,
so what a digit is worth, and every carry and reduction, is learned. That also
sidesteps a hard numeric wall: at Hard's modulus scale `x^2` runs past the 53-bit
exact-integer... | {
"id": "0a016c5a-9549-48ca-ad82-d6df76d47cb4",
"created_at": "2026-08-14 17:25:32.436454+00:00",
"db_md5": "1efadda9e01ae84ec2857432a666c542",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "bde4cc48-83b3-4bed-931a-2710a54dba16",
"tier": "easy",
"dataset_id": "e5",
"status": "su... | {
"score": {
"mean_loss": 2.22281714974049,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004166666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.222699587521532,
"example_count": 600,
... |
0a020499-6013-41e2-b163-2de1e6726861 | easy | rougee | 2026-08-28 03:46:59.207024+00:00 | failed | 2cb33a16629d3825441675f0e740d99aa808f7ebb0899c0ce2d0d05b36431f56 | 7,497 | null | """Official Submittable Champion for One Layer Deeper (Differentiable Algebraic Recurrence).
End-to-End Differentiable Recurrent Architecture:
- Input Representation: Dynamic Token Embedding + Positional Embedding
- Differentiable Residue Extraction: Algebraic position-weighted digit accumulation for N, X, T
- Unrolle... | {
"id": "0a020499-6013-41e2-b163-2de1e6726861",
"created_at": "2026-08-28 03:46:59.207024+00:00",
"db_md5": "83740abcfb187c2eee4dfd40a9c234aa",
"submitter": "Rougee",
"github_login": "rougee",
"run_id": "2c65d746-0973-459a-8f04-095edeb9101b",
"tier": "easy",
"dataset_id": "e1",
"status": "failed",
"... | null
|
0a077057-7277-4aca-9992-9c18d69eb700 | easy | YuMingQian1234 | 2026-08-07 05:24:29.621944+00:00 | succeeded | 5c9a2709b7a74dd43780beadc3fcca335904265a90733f42fc62666a3d51b693 | 3,627 | 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": "0a077057-7277-4aca-9992-9c18d69eb700",
"created_at": "2026-08-07 05:24:29.621944+00:00",
"db_md5": "849aa965c6835feaccf588fa6646dd83",
"submitter": "Mingqian Yu",
"github_login": "YuMingQian1234",
"run_id": "5c03ea2e-1765-47bd-9f49-c43d0faba851",
"tier": "easy",
"dataset_id": "e1",
"status": ... | {
"score": {
"mean_loss": 3.162866473197937,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.006666666828095913
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.208724021911621,
"example_count": 100,... |
0a158166-8362-437f-9ac9-e75889786d1f | easy | TZZheng | 2026-08-06 17:47:21.181333+00:00 | succeeded | 93c9e4797536e8ec3ca543cd8e5bdc1805adcaf1e845b679387d31cd6067c345 | 7,293 | 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": "0a158166-8362-437f-9ac9-e75889786d1f",
"created_at": "2026-08-06 17:47:21.181333+00:00",
"db_md5": "ae9c28eb029f1bca83409738f9df574d",
"submitter": "TZZheng",
"github_login": "TZZheng",
"run_id": "1507be55-5df1-4527-9a63-30ffa3a34488",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.992736339569092,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.06666666455566883
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.034153461456299,
"example_count": 100,
... |
0a163ddd-22ef-4269-9b23-ce0a432b8aec | easy | shan23chen | 2026-08-10 18:46:13.301320+00:00 | succeeded | d178b650993f5827db21214a38046534bd6dcf4e33bb5db15afbcc2b28c75b20 | 40,905 | null | """Dual-root boundary-reset recurrent register Transformer.
One learned transition is applied once per requested outer step. The
transition receives an immutable decimal N register and the current learned
boundary state, but never receives T or a separate copy of the original X.
Within each outer application, scratch... | {
"id": "0a163ddd-22ef-4269-9b23-ce0a432b8aec",
"created_at": "2026-08-10 18:46:13.301320+00:00",
"db_md5": "676c6f81e891f662b490a694cd581956",
"submitter": "Shan Chen",
"github_login": "shan23chen",
"run_id": "be5d5167-0aaa-4eed-9ed6-73480ef611b0",
"tier": "easy",
"dataset_id": "e3",
"status": "succe... | {
"score": {
"mean_loss": 2.260825991630554,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.008749999571591616
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2601075172424316,
"example_count": 800... |
0a1c037d-f2d8-4856-9ff8-cf8ae1c1c42a | easy | gauravmishra | 2026-08-09 22:53:47.058407+00:00 | succeeded | c94c270720868cfb2847fb9284ea0944440ab57d07870981010ef9012376cd2f | 4,268 | 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,
TokenLossBatch,
assert_model_state,
)
D_MODEL = ... | {
"id": "0a1c037d-f2d8-4856-9ff8-cf8ae1c1c42a",
"created_at": "2026-08-09 22:53:47.058407+00:00",
"db_md5": "d7d14efc3f9d1186e18e037d4020d52e",
"submitter": "Gaurav Mishra",
"github_login": "gauravmishra",
"run_id": "2290f8ad-9091-43b7-898a-58863d30cb8e",
"tier": "easy",
"dataset_id": "e5",
"status": ... | {
"score": {
"mean_loss": 2.257241920465522,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.005416666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2843833611150495,
"example_count": 600... |
0a1d6259-c68d-4764-8960-8032b6abe703 | easy | sapient-sapiens | 2026-08-16 16:08:19.693014+00:00 | succeeded | df44a442de0ff45f613add268cab17edb698ee9e5810780f202fcbf0cafcf693 | 22,031 | null | """E6 generalization candidate: width-384 ACT64 shared recurrence.
Architecture changes relative to the hosted ACT max-4 source:
- selectable recurrence depths are 1 through 64;
- canonical cumulative-mass halting genuinely removes halted examples from
later recurrent applications;
- the T-field representation pres... | {
"id": "0a1d6259-c68d-4764-8960-8032b6abe703",
"created_at": "2026-08-16 16:08:19.693014+00:00",
"db_md5": "a93f6080f84fdd40aaf0976ed1210a34",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "2cc92889-77ed-43f3-99cc-2c9f5b9a699c",
"tier": "easy",
"dataset_id": "e6",
"status": "su... | {
"score": {
"mean_loss": 6.298990497850392,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.09707207554900968
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 5.9154582023620605,
"example_count": 60,
... |
0a21e54c-c41f-4948-a797-afdb4412008c | easy | alirezashirvani-jr | 2026-08-31 00:45:34.974067+00:00 | succeeded | d30c15c4b819074a695cfe7f9ab43fd9530cdf8d2f612f55533983ffcf4e7dfb | 22,445 | 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,
)
PERIOD_MINIMUM = 2
PERIOD_MAXIMUM = 96
EXP... | {
"id": "0a21e54c-c41f-4948-a797-afdb4412008c",
"created_at": "2026-08-31 00:45:34.974067+00:00",
"db_md5": "a726edbe0af9de151bd033a13eebcee4",
"submitter": "alirezashirvani-jr",
"github_login": "alirezashirvani-jr",
"run_id": "3c9202ff-18c1-46ca-9200-3652a882f4fb",
"tier": "easy",
"dataset_id": "e2",
... | {
"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": 300,
"exact_accuracy": 1.0,
"... |
0a2289b4-d391-4deb-ac16-3a7a2f3cc5f9 | easy | shan23chen | 2026-08-08 01:56:13.492813+00:00 | succeeded | a26c3067dbc704d97894a935eae5a2d03ba19623f411e75ce9e8496f1dabec83 | 20,626 | null | """Generic recurrent Transformer with normalized numeric and CRF objectives.
The model is deliberately task-generic: it learns a transition over prompt and
workspace tokens instead of embedding the public modular-squaring algorithm.
A learned linear-chain CRF models dependencies among adjacent output digits;
its endpo... | {
"id": "0a2289b4-d391-4deb-ac16-3a7a2f3cc5f9",
"created_at": "2026-08-08 01:56:13.492813+00:00",
"db_md5": "fed02b469c0a393adff4107371773cec",
"submitter": "Shan Chen",
"github_login": "shan23chen",
"run_id": "62337ff0-af31-4068-8151-caa1796597a6",
"tier": "easy",
"dataset_id": "e5",
"status": "succe... | {
"score": {
"mean_loss": 4.074760913848877,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.009583333507180214
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.010627269744873,
"example_count": 600,... |
0a24289e-4ae3-4224-a26d-1fd2b6fa076f | easy | bernettorlando | 2026-08-04 02:30:42.755375+00:00 | succeeded | 16636c7402baae3bebe461ca42bd79d9a803c4bfb6a7a58af31795253ae9cc09 | 5,120 | null | """Beginner-friendly recurrent Transformer submission.
The same learned block is applied several times, so increasing latent depth does
not increase the parameter count. Change only the constants in the
"Experiment knobs" section when you begin experimenting.
"""
from __future__ import annotations
import torch
impo... | {
"id": "0a24289e-4ae3-4224-a26d-1fd2b6fa076f",
"created_at": "2026-08-04 02:30:42.755375+00:00",
"db_md5": "d6f202bab274a8df9141b46b8229219a",
"submitter": "bernettorlando",
"github_login": "bernettorlando",
"run_id": "84a34066-a954-4561-98ae-0cfdaad7d196",
"tier": "easy",
"dataset_id": "e1",
"status... | {
"score": {
"mean_loss": 2.2151116132736206,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0833333320915699
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1771271228790283,
"example_count": 100,... |
0a2c82d1-d14e-49f6-954f-a1dc2bced708 | easy | Laz4rz | 2026-08-07 21:11:25.422253+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": "0a2c82d1-d14e-49f6-954f-a1dc2bced708",
"created_at": "2026-08-07 21:11:25.422253+00:00",
"db_md5": "eaac69fe345f82973fcf392c504a40ff",
"submitter": "Laz4rz",
"github_login": "Laz4rz",
"run_id": "57c9fa21-697d-4624-b0d5-b261b3146da7",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded",
... | {
"score": {
"mean_loss": 3.010505795478821,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0033333334140479565
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.979322671890259,
"example_count": 100... |
0a2f5ea3-2286-440d-8cc4-70e27146a9da | easy | Maxwells-Demons | 2026-08-10 14:32:56.120492+00:00 | succeeded | ce6c977a8aeae56474506410caa9eff071427cd918d10fb54116570c1dde82b2 | 12,508 | null | """v47: the transition as one wide MLP over the digit state and the modulus.
Two measurements drive this shape.
1. The tied attention block (v43-v46) tops out around mean 0.75-0.79 on e1 while
the older flat-MLP transition reached 0.87 on the same data. Attention mixes
slots only through a low-rank path; a dig... | {
"id": "0a2f5ea3-2286-440d-8cc4-70e27146a9da",
"created_at": "2026-08-10 14:32:56.120492+00:00",
"db_md5": "ba4afeef801d311d6cde2f24450e5544",
"submitter": "Maxwells_Ayakashi",
"github_login": "Maxwells-Demons",
"run_id": "a3c16a41-3a86-4f95-8c0e-9d3579d6f454",
"tier": "easy",
"dataset_id": "e2",
"st... | {
"score": {
"mean_loss": 0.15145760774612427,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.9666666984558105
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.17546650767326355,
"example_count": 30... |
0a37c200-ab65-47ae-88b1-fa75e5b84c63 | easy | ipsa-loquitur | 2026-08-26 05:04:08.817024+00:00 | succeeded | 33d10841c5994e90ecbf90aedc455295d9729d9249b1e09b3f302a3f0d184f4d | 15,929 | null | """Standalone NetF-V1 submission candidate for One Layer Deeper.
The learned rule is shared across bit positions, modular-square iterations,
moduli, and tape widths. During training, exact double-and-add states provide
dense one-step supervision. During evaluation, the model receives no state
corrections and rolls out... | {
"id": "0a37c200-ab65-47ae-88b1-fa75e5b84c63",
"created_at": "2026-08-26 05:04:08.817024+00:00",
"db_md5": "61f8c99d46678bc429bca0f01d6f3362",
"submitter": "ipsa-loquitur",
"github_login": "ipsa-loquitur",
"run_id": "1fdf18b0-58a4-4333-bdce-e716880eac27",
"tier": "easy",
"dataset_id": "e4",
"status":... | {
"score": {
"mean_loss": 7.688756516714671,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.6241666662693024
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 10.020926833330961,
"example_count": 1200,... |
0a3aa20a-ae2e-4d39-8e75-0e04fe167982 | easy | msnow0 | 2026-08-08 11:31:29.280826+00:00 | succeeded | 368feb512e5bd7cd9e6219b16e14f82e895a61921d91c83a951839299b8debb4 | 17,246 | null | """One Layer Deeper submission — mechanism 1 (see BRIEF.md §6, research/RULES.md).
Weight-tied recurrent operator over a discrete token tape. One outer iteration per
squaring step (loop count conditioned on the prompt's T, which is explicitly allowed);
between outer steps the latent is snapped back onto token embeddin... | {
"id": "0a3aa20a-ae2e-4d39-8e75-0e04fe167982",
"created_at": "2026-08-08 11:31:29.280826+00:00",
"db_md5": "cc093654a8ff8add7ae3607600f4ffd0",
"submitter": "msnow0",
"github_login": "msnow0",
"run_id": "97be3bf2-9341-47c5-9d6d-e81c07ecf908",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded",
... | {
"score": {
"mean_loss": 2.938732385635376,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.016666666604578495
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.5982303619384766,
"example_count": 100... |
0a3ee973-8fde-4d72-85b3-44621b914b15 | easy | viridale | 2026-08-28 13:36:00.488882+00:00 | succeeded | e8748267744a1e2dff48bfddd20aa7ff70117c88c4519fef5bc4645b5847c8d1 | 30,447 | null | """Wide-slab plus all-medium-cube MDL quadratic program learner.
hypothesis: v983 discovers x^2+117 without T1 and certifies four official identities,
but its union of three slabs allows only one coefficient outside +/-4. Add a separate
all-medium cube with a,b,c each in [-64,64]. This expands the plausible ti... | {
"id": "0a3ee973-8fde-4d72-85b3-44621b914b15",
"created_at": "2026-08-28 13:36:00.488882+00:00",
"db_md5": "96c30621f6eabaf2c390014f91682fcf",
"submitter": "priormancer",
"github_login": "viridale",
"run_id": "733602ad-670e-4c66-b695-fa12f1400325",
"tier": "easy",
"dataset_id": "e3",
"status": "succe... | {
"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": 800,
"exact_accuracy": 1.0,
"... |
0a402d26-c295-4b1f-8512-1a0f638ac736 | easy | DDanlov | 2026-08-14 20:16:23.431114+00:00 | succeeded | ed3b2d23cfb1581e9dc4bee76ac8cace96e1db2fd2fbafe1132e7dafda93662a | 21,983 | 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": "0a402d26-c295-4b1f-8512-1a0f638ac736",
"created_at": "2026-08-14 20:16:23.431114+00:00",
"db_md5": "18f5a2badb0aa94b30c280ca7f8adef4",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "f32506fe-0afe-45b7-80dc-24d47930be8f",
"tier": "easy",
"dataset_id": "e3",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.40039967752954,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.018125000000000002
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.4019566471956835,
"example_count": 800,... |
0a40dc9d-3403-4b27-a58f-4d3c8bf0c7bc | easy | c-nnr | 2026-08-04 12:35:58.967370+00:00 | succeeded | c9b13b919d992f874883990837fe64b45f9e8f2df3b5679ca18a8fc8ff3bc923 | 5,686 | null | """One Layer Deeper E2 round-two experiment: Aggressive learning-rate boundary; detects whether Easy remains under-optimised."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import ModelSpec, OptimizerBundle, OptimizerSpec, Submission, as... | {
"id": "0a40dc9d-3403-4b27-a58f-4d3c8bf0c7bc",
"created_at": "2026-08-04 12:35:58.967370+00:00",
"db_md5": "5a05d28be9e96e8b83880ca172688430",
"submitter": "c-nnr",
"github_login": "c-nnr",
"run_id": "97d9d0fd-7d93-4962-9747-90ccb9ab86ad",
"tier": "easy",
"dataset_id": "e2",
"status": "succeeded",
... | {
"score": {
"mean_loss": 4.24117112159729,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.00541666685603559
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.166479110717773,
"example_count": 300,
... |
0a4409dc-bfaf-4f25-9d08-f1f1b81c796f | easy | shirvani-jr | 2026-09-01 03:58:37.794888+00:00 | succeeded | e4f9cc5ad9870845884601fdfbe39c2ec7c0b4ec2871a6dd672ebc3a13b9a77b | 44,274 | null | from __future__ import annotations
import math
import time
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from torch.utils.checkpoint import checkpoint
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
assert_model_state,... | {
"id": "0a4409dc-bfaf-4f25-9d08-f1f1b81c796f",
"created_at": "2026-09-01 03:58:37.794888+00:00",
"db_md5": "97deb97010056825b0e1fa4a8d743d0d",
"submitter": "Ali",
"github_login": "shirvani-jr",
"run_id": "5f3a628c-99ae-470b-81ce-b6293711ea4a",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 0.9044105638120625,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.9512500254313151
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.9417040348052979,
"example_count": 600,... |
0a4dea6b-b55b-4a06-99c2-b646d83513bc | easy | yashkant | 2026-08-12 03:47:32.998424+00:00 | succeeded | 123eb09e532563114933f204bd31d4cd575dc16247a1314f6c416f4b43f65284 | 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": "0a4dea6b-b55b-4a06-99c2-b646d83513bc",
"created_at": "2026-08-12 03:47:32.998424+00:00",
"db_md5": "12d7d78462e0b64eb6aa7b4972ba3c64",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "75d21ae9-5389-48ff-b49b-4a7e195c2438",
"tier": "easy",
"dataset_id": "e2",
"status": "succeed... | {
"score": {
"mean_loss": 3.5260958075523376,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.1958333421498537
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 5.101095676422119,
"example_count": 300,
... |
0a523227-98d3-4246-8e87-686b637ee3b1 | easy | alirezashirvani-jr | 2026-09-01 01:14:15.373497+00:00 | succeeded | 040b8fb0f1294ce3d51e2a60a1804868a7444703573766d42ab3eeaf61acf75e | 36,899 | 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,
)
MLP_EX... | {
"id": "0a523227-98d3-4246-8e87-686b637ee3b1",
"created_at": "2026-09-01 01:14:15.373497+00:00",
"db_md5": "6e273f136bb17e8b0449146e17366a55",
"submitter": "alirezashirvani-jr",
"github_login": "alirezashirvani-jr",
"run_id": "202878ac-f0f9-4de3-b0d3-345e97836547",
"tier": "easy",
"dataset_id": "e5",
... | {
"score": {
"mean_loss": 2.7980498658107207,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.024166666883975268
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.8460934162139893,
"example_count": 60... |
0a54d12f-87fe-43a7-95f8-8740f3ad5a0c | easy | yashkant | 2026-08-14 15:31:53.698796+00:00 | succeeded | b16837e0c672688c0ed3c5da54128810aaf93fdd210f5212f585458e89ce4742 | 34,010 | null | """Round1795 parsed-T conditional mean-free Dykstra corrections.
Both arms use the strongest Round1792 AdamW5/EMA.995 W8/D2 carrier with
strength-2 contrastive set codes and eagerly compute the same raw and
mean-free correction branches. The candidate keeps Round1792's mean-free
corrections only for public T=2 and se... | {
"id": "0a54d12f-87fe-43a7-95f8-8740f3ad5a0c",
"created_at": "2026-08-14 15:31:53.698796+00:00",
"db_md5": "ee5e787a5c02ccdb8d94b8fb05fe6f02",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "d8871b96-2bf1-4bcb-8bdb-1add1bd81995",
"tier": "easy",
"dataset_id": "e3",
"status": "succeed... | {
"score": {
"mean_loss": 2.1199064354185717,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.01625000007916242
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.119708220197415,
"example_count": 800,... |
0a609782-8d7c-4bb1-9d25-30f2e28660e3 | easy | liam-gb | 2026-08-16 23:06:17.531187+00:00 | succeeded | 82632945ab7e0c87d203dfc4e2f107a9e7f49bf180d7e486a194d56b896646c1 | 14,535 | null | """charblock: a from-scratch body whose per-prime step has a known closed-form
solution, trained in the regime where that solution is documented to be found
from random initialisation.
Five ingredients, declared as one ARCHITECTURE+REGIME bundle:
1. Per-prime two-layer blocks with a QUADRATIC activation. For each ban... | {
"id": "0a609782-8d7c-4bb1-9d25-30f2e28660e3",
"created_at": "2026-08-16 23:06:17.531187+00:00",
"db_md5": "a79f5e7cd40035de9d3d3adb1e10b7b5",
"submitter": "liam-gb",
"github_login": "liam-gb",
"run_id": "edbefbb8-8b89-446d-8ab2-c94261f7fefa",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.363173723220825,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.3883333345875144
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.57509446144104,
"example_count": 100,
... |
0a62972c-b100-4e7d-9882-3a75b1dfa487 | easy | khushidahi | 2026-08-16 02:50:55.297907+00:00 | succeeded | 54aacbfa8e8d9eb32bda557d84d9fe5384b716b309e3f09c23901aca55745d53 | 22,535 | null | """R91: ordered-modulus conditioned transition.
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 transitio... | {
"id": "0a62972c-b100-4e7d-9882-3a75b1dfa487",
"created_at": "2026-08-16 02:50:55.297907+00:00",
"db_md5": "357458efa2cfc534688bfe810d031884",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "62f07b6d-dac2-4be9-9c73-0112dc4c3490",
"tier": "easy",
"dataset_id": "e5",
"status": "succ... | {
"score": {
"mean_loss": 2.6460492351509544,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0275
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.657970845432025,
"example_count": 600,
"e... |
0a663c88-ca45-46b5-980c-19f9de0fc3af | easy | im-a-king-but-im-far-from-a-saint | 2026-08-08 14:58:07.854965+00:00 | succeeded | 790da41d05de3fb8ca0cc2e69de83f4caa1f66bdd67ff2f8bbaa75dbd1d1cf07 | 5,138 | null | """Weight-tied recurrent (Universal-Transformer-style) baseline.
Applies ONE shared transformer block NUM_LOOPS times instead of stacking
distinct layers. Re-injecting the original embedding at every iteration
keeps the recurrence anchored to the problem statement, which is what lets
this kind of network extrapolate p... | {
"id": "0a663c88-ca45-46b5-980c-19f9de0fc3af",
"created_at": "2026-08-08 14:58:07.854965+00:00",
"db_md5": "1cd5cc79d771306848f206479e874c0f",
"submitter": "im-a-king-but-im-far-from-a-saint",
"github_login": "im-a-king-but-im-far-from-a-saint",
"run_id": "a9e7cc56-a141-4adf-a8f9-5bfb003c3b5d",
"tier": "... | {
"score": {
"mean_loss": 2.5597325563430786,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.008333333302289248
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.9559218883514404,
"example_count": 10... |
0a66a960-d89f-487b-aabf-dd8e865939c6 | easy | yashkant | 2026-08-10 21:32:03.207506+00:00 | succeeded | 5009134053ca85753760fce70a90333e1b374205628be5201f1e7dc4459fe898 | 22,265 | null | """Standalone compact-bigram hash, AdamW LR3.75e-3, EMA 0.997."""
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": "0a66a960-d89f-487b-aabf-dd8e865939c6",
"created_at": "2026-08-10 21:32:03.207506+00:00",
"db_md5": "d928b459701ca69468b300c96bb056f8",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "84b15f33-1198-49b4-af36-15eecced7284",
"tier": "easy",
"dataset_id": "e5",
"status": "succeed... | {
"score": {
"mean_loss": 2.16515671575146,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.005416666691501935
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1671581899104098,
"example_count": 600,... |
0a838bd3-4709-4f1e-b1d8-6e217abb975e | easy | viridale | 2026-08-06 12:02:30.381010+00:00 | succeeded | 6bd5f36a4fa460688d8bb6f4bac1dc3aa9caaf55d30c2435cc9fff07bf593455 | 10,926 | null | """Encoder-first training followed by a fresh frozen-encoder decoder.
hypothesis: useful endpoint structure can emerge in the recurrent encoder before its
jointly trained tied readout can use it; freezing that representation and training
a previously untouched generic decoder should remove readout interference... | {
"id": "0a838bd3-4709-4f1e-b1d8-6e217abb975e",
"created_at": "2026-08-06 12:02:30.381010+00:00",
"db_md5": "4ee09d9d18f53ae4751b04e01cd27921",
"submitter": "priormancer",
"github_login": "viridale",
"run_id": "1771c869-0c65-4971-bc62-705c79be075f",
"tier": "easy",
"dataset_id": "e1",
"status": "succe... | {
"score": {
"mean_loss": 2.184759736061096,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0833333320915699
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2082674503326416,
"example_count": 100,
... |
0a870b72-5cd0-4609-8777-073216fab64a | easy | lpbb | 2026-08-18 09:42:35.229884+00:00 | succeeded | 8c67d267e8fb237789964945c760bf7827d7839af22146990057c2a2e6ace7f7 | 4,457 | null | """S0 — control: same shape as S1 but 24 untied layers (fixed depth, no weight tying).
Measures the 'tying tax'; by construction its depth cannot follow T."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
Optimi... | {
"id": "0a870b72-5cd0-4609-8777-073216fab64a",
"created_at": "2026-08-18 09:42:35.229884+00:00",
"db_md5": "c76e873aa9eb43af19a94257b004713c",
"submitter": "Lpbb",
"github_login": "lpbb",
"run_id": "67ccbd6b-dc23-4e72-8e9d-6b9eb0c610e7",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded",
"s... | {
"score": {
"mean_loss": 7.596846103668213,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.031666665337979794
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 7.063924789428711,
"example_count": 100,... |
0a8baf89-a99e-4ef5-9188-a2ceafe51a7e | easy | Yalyenea | 2026-08-12 13:00:43.134771+00:00 | succeeded | c737baf40609147da35aa2a880cc07185bc0ce37caddb72d9732d584ac0f3ac2 | 11,810 | null | """Class-gated recurrence with dropout, label smoothing, 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": "0a8baf89-a99e-4ef5-9188-a2ceafe51a7e",
"created_at": "2026-08-12 13:00:43.134771+00:00",
"db_md5": "3f55aff46bb6fc7e646ff5425f081b90",
"submitter": "yfff",
"github_login": "Yalyenea",
"run_id": "7567b17c-ba7b-435a-98d9-bb865c03e548",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded",
... | {
"score": {
"mean_loss": 2.8871380990517164,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.007916666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.5658752458512515,
"example_count": 60... |
0a8ed77b-da19-4d2a-9a44-e7676f4f4a8e | easy | ehonig | 2026-08-20 04:12:15.309771+00:00 | succeeded | 634c6b23b43ea904fdad6c235da40f8112a1b5e5e6976b651343a466c9910f24 | 22,766 | 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": "0a8ed77b-da19-4d2a-9a44-e7676f4f4a8e",
"created_at": "2026-08-20 04:12:15.309771+00:00",
"db_md5": "0e0b7c85c878bd05d7143dfb80107df8",
"submitter": "Edouardo Honig",
"github_login": "ehonig",
"run_id": "a8a79381-27f8-4807-a4ac-34f1d37e68c5",
"tier": "easy",
"dataset_id": "e1",
"status": "succ... | {
"score": {
"mean_loss": 2.1679792404174805,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0833333320915699
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1931443214416504,
"example_count": 100,... |
0a99dfa1-534f-489d-bf19-edaf4b3deb22 | easy | gauravmishra | 2026-08-08 00:09:51.516796+00:00 | succeeded | d74892c24149a43ba0cdffb61184aad8bf53fa05abdc4d223f4d559049fc2bc8 | 3,657 | 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": "0a99dfa1-534f-489d-bf19-edaf4b3deb22",
"created_at": "2026-08-08 00:09:51.516796+00:00",
"db_md5": "85d8ec03e0b3a6a9bd08ce2324f23cdd",
"submitter": "Gaurav Mishra",
"github_login": "gauravmishra",
"run_id": "3c71c984-c6f8-408c-a9f7-db15f114b86e",
"tier": "easy",
"dataset_id": "e5",
"status": ... | {
"score": {
"mean_loss": 2.2352964306964402,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.005000000037252903
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.276047017007665,
"example_count": 600... |
0aa50ce6-793b-4e93-b810-b2295551e0b8 | easy | sapient-sapiens | 2026-08-04 16:10:30.756506+00:00 | succeeded | 538e90d605a78b31f2e80ba6632a9da364afd55cabb4c5b7bce5a7dcd1e40699 | 18,014 | null | """Weight-tied recurrent-depth Transformer with progressive + hold losses.
Phase 2: prelude / recalled tied core / coda, differentiable hold gate,
fp32 core under disabled autocast, random train depth, K=72 at eval.
Extra depth readouts travel through the auxiliary return value into
token_training_loss. Wall-clock Ada... | {
"id": "0aa50ce6-793b-4e93-b810-b2295551e0b8",
"created_at": "2026-08-04 16:10:30.756506+00:00",
"db_md5": "b5fbc952bcd67108c609ff83e3186b12",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "0a6fd463-f4c6-4e6e-98c7-7d0ffc39bcdd",
"tier": "easy",
"dataset_id": "e1",
"status": "su... | {
"score": {
"mean_loss": 2.0110684264211804,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.03000000017384688
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.9319810604871208,
"example_count": 100... |
0ab3b0ad-81a7-4727-b85a-bad5631b3390 | easy | Yalyenea | 2026-08-12 07:21:55.607738+00:00 | succeeded | 244a0ef4e9b84016e991b775c4c1c85e037ef2062d96cf6a661f76dc99a92667 | 10,753 | null | """Single-block decimal macro-transition with one-step supervision."""
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": "0ab3b0ad-81a7-4727-b85a-bad5631b3390",
"created_at": "2026-08-12 07:21:55.607738+00:00",
"db_md5": "41de07636f4641cfba2c3ca17c641879",
"submitter": "yfff",
"github_login": "Yalyenea",
"run_id": "51524bf2-b8d9-49c5-9122-ff3a94b394e8",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded",
... | {
"score": {
"mean_loss": 4.966128853907133,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.012500000012417634
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.752526651048874,
"example_count": 600,... |
0ab3eee5-4e62-4d2b-94e5-eff80f6f4d24 | easy | 0Chris5R | 2026-08-07 09:30:07.080877+00:00 | succeeded | a865fd58467b787591d8b4bb8fb59512029d483388b955377126621d7bb2bee1 | 14,294 | null | """Task-agnostic token-space function composer with exact structural T loops."""
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,
Tok... | {
"id": "0ab3eee5-4e62-4d2b-94e5-eff80f6f4d24",
"created_at": "2026-08-07 09:30:07.080877+00:00",
"db_md5": "8c6a7ba4ca67a3f137b27ca53120ca7a",
"submitter": "Chris ",
"github_login": "0Chris5R",
"run_id": "da737e42-0e60-4d67-b539-0aad55087833",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 8.281500086217815,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.007083333333333334
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 8.185522006766144,
"example_count": 600,... |
0ab9ff21-4d3a-4af4-95ee-009b91adb234 | easy | 0Chris5R | 2026-08-02 16:49:17.831199+00:00 | succeeded | 0e1b2872fcfaff5cffc26c03d76448e6b5bc5596589c1e42ad0410a0205d608e | 24,605 | null | """Shared-digit Neural GPU with late full-objective EMA evaluation.
The model aligns decimal digits by significance, forms a learned low-rank
second-order feature map of the current state, and evolves a two-dimensional
workspace with the original convolutional-GRU equations. Each arithmetic
transition applies the sam... | {
"id": "0ab9ff21-4d3a-4af4-95ee-009b91adb234",
"created_at": "2026-08-02 16:49:17.831199+00:00",
"db_md5": "cd93dd4f81dbb7b0b61c5d6ddb59b333",
"submitter": "Chris ",
"github_login": "0Chris5R",
"run_id": "ea4e2eef-e5fe-404a-bdd5-aed6563ca7a8",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.517346321862235,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.035416666703919575
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.3919387473119214,
"example_count": 600... |
0abe6159-b21b-465e-ac71-6a7665b32440 | easy | chad-atexpedient | 2026-08-14 16:26:57.889699+00:00 | succeeded | 5a15abaed28c208af11091f66b7766cf2d9f81ed0a3ff64a5c8fece2034bcd7c | 17,596 | null | """X41: Hybrid delta with higher cross-N weight (0.05/0.15) for OOD-N."""
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_st... | {
"id": "0abe6159-b21b-465e-ac71-6a7665b32440",
"created_at": "2026-08-14 16:26:57.889699+00:00",
"db_md5": "13065d91f4e22f1fa0f9cb8f21485f5f",
"submitter": "chad-atexpedient",
"github_login": "chad-atexpedient",
"run_id": "4a8a82ca-931d-4060-9922-3f0e5ee351ad",
"tier": "easy",
"dataset_id": "e5",
"st... | {
"score": {
"mean_loss": 2.305890669393122,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.013333333345750969
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.3608780944293923,
"example_count": 600... |
0ac30d0b-3a78-4b88-a327-106ba2071574 | easy | repst | 2026-08-15 00:30:11.035806+00:00 | failed | db0e60a1617170fa3249a06eeb2cfbf8fa1c581f039c293132c12e4a0ee1e26d | 28,148 | null | """r260: exact semantic closure after behavioral compilation.
This standalone r246 successor factors one learned schedule bit from the typed
instruction slots. The stationary choice exactly preserves r246; the bootstrap
choice enables slot 2 on even-numbered phases while slot 1 remains recurrent. The
machine still ex... | {
"id": "0ac30d0b-3a78-4b88-a327-106ba2071574",
"created_at": "2026-08-15 00:30:11.035806+00:00",
"db_md5": "35dc2302591bb4c8ea4ab2be66d36037",
"submitter": "Ryan Epstein",
"github_login": "repst",
"run_id": "0ea7f1a1-de54-4e5b-9d07-4bc3cc88f2fd",
"tier": "easy",
"dataset_id": "e1",
"status": "failed"... | null
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.