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 |
|---|---|---|---|---|---|---|---|---|---|---|
4403a47b-a238-48e1-a5a9-5965d6bec332 | easy | liam-gb | 2026-08-13 13:09:51.895501+00:00 | succeeded | c98994ef0a17cbbc563e971c85fd060c426e82414551fdb5ce599d6575496cbc | 15,106 | null | """rns_pm_gather_ct_widet_pt: widet + per-prime temperature (16-prime bank).
The wide readout is what makes long answers reachable; the trim is what pays
for it, by not running loop iterations whose update is multiplied by zero.
Base docstring:
Single change vs rns_pm_gather_ct: the readout carries one
distance-from... | {
"id": "4403a47b-a238-48e1-a5a9-5965d6bec332",
"created_at": "2026-08-13 13:09:51.895501+00:00",
"db_md5": "f909714a31387e09614b38df91ccc3f4",
"submitter": "liam-gb",
"github_login": "liam-gb",
"run_id": "cb460fff-e5bd-4000-a06e-2647145f08e7",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 5.185023605823517,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.486666664481163
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 9.129050254821777,
"example_count": 100,
... |
44047784-a0f0-46d0-91e2-090311e1790e | easy | chad-atexpedient | 2026-08-05 16:39:05.106361+00:00 | succeeded | c622621bf98fe502d08550d5fee1af80b93353345a7814ff267ed720c1800e8e | 13,112 | null | """R032: three shared learned macrosteps with four microsteps per transition."""
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,
)
LANE... | {
"id": "44047784-a0f0-46d0-91e2-090311e1790e",
"created_at": "2026-08-05 16:39:05.106361+00:00",
"db_md5": "5f0bd5b60e92447d3a25882b5bfb0b8f",
"submitter": "chad-atexpedient",
"github_login": "chad-atexpedient",
"run_id": "6acc5679-4d67-4129-8661-cb5b1c9fd93a",
"tier": "easy",
"dataset_id": "e5",
"st... | {
"score": {
"mean_loss": 2.190476262069291,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.005833333333333334
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.188894279334577,
"example_count": 600,... |
440e37c2-32e7-4c39-b3de-3b6813367c70 | easy | yunjiangster | 2026-08-26 16:10:57.146161+00:00 | failed | 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": "440e37c2-32e7-4c39-b3de-3b6813367c70",
"created_at": "2026-08-26 16:10:57.146161+00:00",
"db_md5": "eaac69fe345f82973fcf392c504a40ff",
"submitter": "Yunjiang Jiang",
"github_login": "yunjiangster",
"run_id": "80bd86a6-05e9-4ab2-bb9b-0bc32e61c464",
"tier": "easy",
"dataset_id": "e6",
"status":... | null
|
440f0f71-71ba-44e3-9e15-11f31278cd6a | easy | EyimofeA | 2026-08-15 18:50:00.820107+00:00 | succeeded | 4c82e38a57035df521a4ad8e97dff35c3a9ce1a2bc1335b5c4f4c2c938ac7949 | 7,926 | null | """Exact binary representation + learned squarer + learned prefix reducer.
All trainable weights start random and receive only evaluator final-label loss.
The decimal-to-binary conversion is representation preprocessing, not a task
solver. Squaring, reduction, recurrence, and decimal decoding are learned.
"""
from __f... | {
"id": "440f0f71-71ba-44e3-9e15-11f31278cd6a",
"created_at": "2026-08-15 18:50:00.820107+00:00",
"db_md5": "06ade301b266e1a833f4af5ee8b138f8",
"submitter": "mof",
"github_login": "EyimofeA",
"run_id": "c1c7564f-d39f-421c-81ca-313fcdc625e6",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded",
... | {
"score": {
"mean_loss": 2.303283436454353,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0054166666790843015
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.308427209143147,
"example_count": 600... |
440fe718-779b-438b-a6ce-b9cc1290f68a | easy | erdavis0 | 2026-08-23 16:01:06.483661+00:00 | succeeded | a088f356590cfba5ad21d2d47f6b52ef6ea3fd6c700f05b01303b3ce92032250 | 23,907 | 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": "440fe718-779b-438b-a6ce-b9cc1290f68a",
"created_at": "2026-08-23 16:01:06.483661+00:00",
"db_md5": "3e958839de80efbf59a22b9e0b2a18fd",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "58825005-2cb0-4ca0-9c0f-e75709aeecf9",
"tier": "easy",
"dataset_id": "e7",
"status": "succeeded",... | {
"score": {
"mean_loss": 7.275882720947266,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.06898395903408527
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 8.132415771484375,
"example_count": 85,
... |
4412cb20-68fa-4b14-9fc1-1910855ae51c | easy | anirudh-chakravarthy | 2026-08-12 03:32:32.814341+00:00 | succeeded | ae3c9f1b6e216c52767c4f6f68c74572a84e5789dfdf53a9383edf4274c3d28b | 14,824 | null | """A learned state transition applied once per requested time step."""
from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
asser... | {
"id": "4412cb20-68fa-4b14-9fc1-1910855ae51c",
"created_at": "2026-08-12 03:32:32.814341+00:00",
"db_md5": "cc228ac60ecfce164c7e6ae22b808dd8",
"submitter": "Anirudh S Chakravarthy",
"github_login": "anirudh-chakravarthy",
"run_id": "e8670ef9-b2d6-4f09-9574-b780a2cbe2b2",
"tier": "easy",
"dataset_id": "... | {
"score": {
"mean_loss": 3.081318774692238,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.009375000060535968
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.324697253580699,
"example_count": 800,... |
441342a0-38cc-4011-aadb-38b8d01e2a66 | easy | blake-camp-surge | 2026-08-05 18:41:26.026483+00:00 | succeeded | 7ee22a7badaa94dfe4426db6155721702558ced8c9523e42fe82f1065abf5ca1 | 10,387 | null | """Vector-state neural cellular automaton over attention edges."""
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,
)
TOKEN... | {
"id": "441342a0-38cc-4011-aadb-38b8d01e2a66",
"created_at": "2026-08-05 18:41:26.026483+00:00",
"db_md5": "78b65d36687cfcd81a2c228873033dd2",
"submitter": "Blake Camp",
"github_login": "blake-camp-surge",
"run_id": "7900f231-6605-43d5-bf29-2f12ed983abb",
"tier": "easy",
"dataset_id": "e3",
"status":... | {
"score": {
"mean_loss": 4.754902361435186,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0075
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.885876608235925,
"example_count": 800,
"ex... |
4414ffb9-1350-42fd-929f-836061b81e44 | easy | alirezashirvani-jr | 2026-08-30 21:43:01.078975+00:00 | failed | b7f074324abccefa390459edb027c617899eb92d88f6373fe8c4d4fed50e6fc2 | 27,448 | 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_WIDTH = 64
ATTENTION_HEADS = ... | {
"id": "4414ffb9-1350-42fd-929f-836061b81e44",
"created_at": "2026-08-30 21:43:01.078975+00:00",
"db_md5": "b152aad48c714139d18f722dc39fdeb4",
"submitter": "alirezashirvani-jr",
"github_login": "alirezashirvani-jr",
"run_id": "3359ead6-325f-4756-8b62-ba95dac7c7bb",
"tier": "easy",
"dataset_id": "e3",
... | null
|
4424e0ff-c9ad-457f-b1b2-a62cec5c1875 | easy | erdavis0 | 2026-08-27 01:06:13.085642+00:00 | succeeded | dea5a50d1a7e1542b606c505bb2e064069826b4fcc486a84827b8cbefc9b781f | 31,293 | null | """Typed-prefix reversible transition B0: ordinary evaluator token CE.
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": "4424e0ff-c9ad-457f-b1b2-a62cec5c1875",
"created_at": "2026-08-27 01:06:13.085642+00:00",
"db_md5": "9a82ec9520ce6d08ad83d5dbf700b0ab",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "f329263a-4ca9-4b94-8ad6-91d5e7c8508d",
"tier": "easy",
"dataset_id": "e8",
"status": "succeeded",... | {
"score": {
"mean_loss": 3.9635770320892334,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.1143493801355362
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.406785011291504,
"example_count": 85,
... |
442d1f95-2540-4c37-86a2-80b93e507a09 | easy | sapient-sapiens | 2026-08-06 20:50:42.481171+00:00 | succeeded | 21092fe5a42e150b35a081d18e5d39657c5749773e18f198f9c7d599c74c6bd6 | 8,863 | null | """Learned Fourier arithmetic model, specialized for one-step generalization.
The network receives generic decimal values and polynomial features, then
learns its own random Fourier basis. It does not implement modular reduction
or contain a modulus-specific lookup table.
"""
from __future__ import annotations
impor... | {
"id": "442d1f95-2540-4c37-86a2-80b93e507a09",
"created_at": "2026-08-06 20:50:42.481171+00:00",
"db_md5": "9528326c6e00b2a700610c1ef988e789",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "318e0111-9abe-44c1-9829-df4920ba882f",
"tier": "easy",
"dataset_id": "e1",
"status": "su... | {
"score": {
"mean_loss": 12.382617473602295,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.19500000681728125
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 15.366748809814453,
"example_count": 100... |
442ed0bd-021d-4866-b24b-7b4720e9ac1b | easy | sapient-sapiens | 2026-08-13 20:13:57.764837+00:00 | succeeded | 64ec8386e7a66727bfc4d8ce04d51e432a8142ad25f6c5a24651cf50c0e930e7 | 5,980 | null | """Paper-style additive-input-injection loop for fixed-T=2 E3."""
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, assert_model_state
WIDTH = 1536
HEADS =... | {
"id": "442ed0bd-021d-4866-b24b-7b4720e9ac1b",
"created_at": "2026-08-13 20:13:57.764837+00:00",
"db_md5": "5cd0c896fdf7b9a9bee586e6523f06bf",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "ea381d75-418c-4407-b26b-1b12ff116fba",
"tier": "easy",
"dataset_id": "e3",
"status": "su... | {
"score": {
"mean_loss": 7.3215460442990175,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.006875000000000001
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 7.501387779081775,
"example_count": 800... |
4433f894-c326-46a6-ad89-5824c8821ebe | easy | DDanlov | 2026-08-24 20:39:13.208180+00:00 | succeeded | 2f7a719305e469d56293e5766e93cd20ea34e6de8ef66fa832cb3d392b21c8d9 | 49,569 | null | """
Unified Pure Skipless Prefix Deliberation Model + Transformer Readout Decoder
Architecture:
- Deliberation Core: dim=1536, num_heads=24, d_ff=6144 (Pure Skipless SO(2)^{D/2} Givens Recurrence)
- Token Layout: Recurrent loop processes ONLY [Prompt (L_p), 4 Pause Registers] (Zero target tokens in loop)
- Recurrent At... | {
"id": "4433f894-c326-46a6-ad89-5824c8821ebe",
"created_at": "2026-08-24 20:39:13.208180+00:00",
"db_md5": "9812e56d3ebc52f5eda6b55938080448",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "c7a90628-246c-4106-930a-db59d57e0b47",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.7161348085154122,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004166666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.746987926585792,
"example_count": 600... |
443caac1-c447-4730-b907-b20a1f44cc54 | easy | DDanlov | 2026-08-26 20:48:15.672660+00:00 | succeeded | 77d161ba15c5ea6f917f98784c157d1161da2b469a2437c62a0cf80e147369c9 | 34,346 | null | import math
import time
from dataclasses import dataclass
from typing import Optional, Tuple, Dict, Any, List, Union
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import Optimizer
try:
from onelayerdeeper.api import Submission, OptimizerBundle, ModelSpec, OptimizerSpec
except ... | {
"id": "443caac1-c447-4730-b907-b20a1f44cc54",
"created_at": "2026-08-26 20:48:15.672660+00:00",
"db_md5": "73d718f272edb4bd083c00c490b917fe",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "78dfde55-6695-444e-8674-ccd499e1d9a2",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.3280942862539544,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004166666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.3258252603590757,
"example_count": 60... |
4441e570-a94a-41e2-be31-dafaa978b820 | easy | sapient-sapiens | 2026-08-17 17:54:31.313997+00:00 | succeeded | 853099ed2d8a675acdec11de42d42228cc6c913667f973274419166c4252761b | 30,225 | null | """Field-local ACT candidate: T-free operator, countable halt, answer-slot readout.
Easy-screen cap is MAX_CHUNKS=8. Raise to 64 before a certification packet or Hard.
SAM is kept. T is never used to compute the answer. Token sequences of N/X/T are
matched in the loss only to route existing training labels onto recurr... | {
"id": "4441e570-a94a-41e2-be31-dafaa978b820",
"created_at": "2026-08-17 17:54:31.313997+00:00",
"db_md5": "1caea6aafb15190f270ce890a5fbc456",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "09591b0b-aab8-48b8-9892-18f58b905584",
"tier": "easy",
"dataset_id": "e6",
"status": "su... | {
"score": {
"mean_loss": 1.9461558673479784,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.023423424292657827
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.9340521097183228,
"example_count": 60... |
4444f130-3e50-4c90-8457-aacf76fabe9f | easy | khushidahi | 2026-08-07 20:22:30.566828+00:00 | succeeded | 12010da0463eac5aa384727c1fe03b6cc434caafceaabac476e17b8829d10f33 | 8,530 | null | """H0 conservative R3 Hard anchor.
This is the existing R3 architecture and AdamW recipe evaluated after exactly
3,000 optimizer steps. It provides the matched control for the boundary-aware
sequence-loss experiment.
"""
from __future__ import annotations
import random
import torch
import torch.nn.functional as F
f... | {
"id": "4444f130-3e50-4c90-8457-aacf76fabe9f",
"created_at": "2026-08-07 20:22:30.566828+00:00",
"db_md5": "ef3bcc9657d3fe0bf37b7e7e61f6d5c5",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "6d85edee-b46a-425c-b45e-dea457fb5bf3",
"tier": "easy",
"dataset_id": "e5",
"status": "succ... | {
"score": {
"mean_loss": 2.1721662282943726,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004583333502523601
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.177706480026245,
"example_count": 600... |
44565bac-5891-4a3b-8d9e-4c649ef93744 | easy | DESU-CLUB | 2026-08-30 21:58:35.173788+00:00 | succeeded | 98055c59daf56e75532f422f6df0b65fdf8d0c16f68b055115e9483894363129 | 15,591 | null | """Discrete-state loop over a place-value tape, trained with orthogonalize-only Muon.
The recurrent state carried between squarings is the *digit string*, not a
residual stream. Each outer iteration reads (N, x) off an LSB-first place-value
tape, emits digit logits for one squaring, hard-argmaxes them (straight-throu... | {
"id": "44565bac-5891-4a3b-8d9e-4c649ef93744",
"created_at": "2026-08-30 21:58:35.173788+00:00",
"db_md5": "d539f6cda57908d480762a3affdd1af5",
"submitter": "Low Keng Hoong, Warren",
"github_login": "DESU-CLUB",
"run_id": "11faf6ad-08d7-4cf4-8da7-835d658b8e71",
"tier": "easy",
"dataset_id": "e6",
"sta... | {
"score": {
"mean_loss": 1.0571774966912728,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.75923426167385
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.5375757217407227,
"example_count": 60,
... |
4457e196-a841-48f3-8df1-cd9a559be0da | easy | ehonig | 2026-08-19 17:24:42.931571+00:00 | succeeded | b4cfda33150c97b4231bd7371c9494adee72070e0d8190e243262ea64b8a78d7 | 18,780 | 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": "4457e196-a841-48f3-8df1-cd9a559be0da",
"created_at": "2026-08-19 17:24:42.931571+00:00",
"db_md5": "e10b4be666283c83533ad07629b30349",
"submitter": "Edouardo Honig",
"github_login": "ehonig",
"run_id": "4298326d-4c78-4e51-bb9b-3441b51f91c4",
"tier": "easy",
"dataset_id": "e1",
"status": "succ... | {
"score": {
"mean_loss": 1.8554800152778625,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.07999999821186066
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.8227862119674683,
"example_count": 100... |
445de3c2-a7b0-4ec4-8a29-4097e7203be1 | easy | alirezashirvani-jr | 2026-08-30 22:23:06.678135+00:00 | succeeded | 5a88ea41463e333fbe1a2deb44699a8c84ff6760cf9b2ed00fed3686c586d018 | 24,989 | 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_WIDTH = 32
ATTENTION_HEADS = ... | {
"id": "445de3c2-a7b0-4ec4-8a29-4097e7203be1",
"created_at": "2026-08-30 22:23:06.678135+00:00",
"db_md5": "c05ac8c34491c0c58c3953ea688ec5ce",
"submitter": "alirezashirvani-jr",
"github_login": "alirezashirvani-jr",
"run_id": "0d77fc0e-b463-427a-a8eb-91a09b62e410",
"tier": "easy",
"dataset_id": "e3",
... | {
"score": {
"mean_loss": 2.2256266496019768,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.00375
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.222715182418727,
"example_count": 800,
"... |
44602903-1b94-4764-8f5c-94a3a7875337 | easy | bk3015 | 2026-08-02 23:22:03.929687+00:00 | succeeded | 105094fc0e5abdd3d44725945e03d41b3a72b414fd0deba3a1524ad22300074b | 11,489 | null | """Attention-free recurrent diagonal convolutional register machine.
This file is self-contained for the One Layer Deeper submission contract. It
uses a decimal-token tape because that is the evaluator's public interface;
field tokens and tape boundaries replace task-specific binary preprocessing.
"""
from __future_... | {
"id": "44602903-1b94-4764-8f5c-94a3a7875337",
"created_at": "2026-08-02 23:22:03.929687+00:00",
"db_md5": "a2a1208f5576b19cf0f7888c4b7e24df",
"submitter": "bk3015",
"github_login": "bk3015",
"run_id": "7d59bd83-cc8d-4d3a-b68c-502f81b6f374",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded",
... | {
"score": {
"mean_loss": 2.6660088300704956,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0833333320915699
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.668095827102661,
"example_count": 100,
... |
4463127a-d5d8-4635-bb59-6810adc7ad2d | easy | shan23chen | 2026-08-08 02:08:49.888411+00:00 | succeeded | eeab35fa7dfd3d15026f7cd071a28128a0d9ed81d6c1c745d13796d73d8a34aa | 20,442 | 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": "4463127a-d5d8-4635-bb59-6810adc7ad2d",
"created_at": "2026-08-08 02:08:49.888411+00:00",
"db_md5": "9d51d5f7973e0d75c55917dc0ff376e1",
"submitter": "Shan Chen",
"github_login": "shan23chen",
"run_id": "68e8477b-f620-4f07-ba3c-2ae30723d613",
"tier": "easy",
"dataset_id": "e5",
"status": "succe... | {
"score": {
"mean_loss": 3.0585159063339233,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.011250000447034836
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.017101287841797,
"example_count": 600... |
446e9b88-2078-4ecf-80b9-1c0fbdad4b54 | easy | benjaminW2025 | 2026-08-17 07:29:44.657939+00:00 | succeeded | 12cb7b1530bf8d6834004a9783a4f1233424799ab4a809e0c5c0c5bdcc0296fc | 3,664 | 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 = 256
NUM_HEADS = 4
... | {
"id": "446e9b88-2078-4ecf-80b9-1c0fbdad4b54",
"created_at": "2026-08-17 07:29:44.657939+00:00",
"db_md5": "07dd19f9c21d332bf663f1cb359860dd",
"submitter": "benjawesome",
"github_login": "benjaminW2025",
"run_id": "5dbe7058-e992-47ee-8dda-98983eeb7905",
"tier": "easy",
"dataset_id": "e5",
"status": "... | {
"score": {
"mean_loss": 2.2800376808494827,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004166666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2841198337452293,
"example_count": 60... |
4477946a-c0ee-4b9c-b4e5-8c2b0fe5074c | easy | AdityaVG13 | 2026-08-13 04:54:07.404716+00:00 | succeeded | 5015b724e0d16ba55f82d8334906e3fd36aaf1ae6e02f2b223316d40674c158d | 49,558 | null | """RADIUS V4: Recovery-Aware Dyadic Inference with Uncertainty Splitting.
One Layer Deeper submission candidate. Every task transition is learned
end-to-end. The deterministic shell only (1) separates marker-delimited numeric
fields, (2) represents decimal digits on a fixed semantic tape, and (3) schedules
learned t... | {
"id": "4477946a-c0ee-4b9c-b4e5-8c2b0fe5074c",
"created_at": "2026-08-13 04:54:07.404716+00:00",
"db_md5": "6fbcc9204e67d1c5080647e4e67c835f",
"submitter": "AdityaG",
"github_login": "AdityaVG13",
"run_id": "d8420d97-5050-4f5c-b365-0bfc616de53e",
"tier": "easy",
"dataset_id": "e1",
"status": "succeed... | {
"score": {
"mean_loss": 1.746880292892456,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.31166667491197586
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.7852340936660767,
"example_count": 100,... |
44801d66-c444-4f50-b8e2-60744325d10b | easy | jordanrubin | 2026-08-01 20:48:30.660125+00:00 | succeeded | 2c6adbc91db77091eaa1fdb53b2b15fdc6e05306608a2440be5e6a2aa2591b8a | 18,183 | null | """Looped bidirectional selective-SSM (Mamba-style), pure PyTorch.
The statefulness bet: input-dependent SSM dynamics along the token axis,
with the block weight-tied and applied NUM_LOOPS times in depth with input
injection — recurrence in depth is where the task's T-iteration lives.
The selective scan is computed i... | {
"id": "44801d66-c444-4f50-b8e2-60744325d10b",
"created_at": "2026-08-01 20:48:30.660125+00:00",
"db_md5": "35fec302dbad6ab1de27b5092ede0b6a",
"submitter": "Jordan Rubin",
"github_login": "jordanrubin",
"run_id": "65a72e5c-e495-4cfd-a2d0-95a87a9c723e",
"tier": "easy",
"dataset_id": "e5",
"status": "s... | {
"score": {
"mean_loss": 4.224352085816231,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004999999888241291
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.327707750647592,
"example_count": 600,... |
4480b680-a1e2-47bb-b70e-ce63b249281d | easy | im-a-king-but-im-far-from-a-saint | 2026-08-08 13:55:50.749468+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": "4480b680-a1e2-47bb-b70e-ce63b249281d",
"created_at": "2026-08-08 13:55:50.749468+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": "739ac1a6-93dc-492a-9adc-f00715ed0097",
"tier": "... | {
"score": {
"mean_loss": 2.3633339405059814,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0033333334140479565
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.679926633834839,
"example_count": 10... |
44865a0b-344c-486c-96fc-fc8f81e1a8ef | easy | DDanlov | 2026-08-15 21:28:14.397626+00:00 | failed | 678841da09b3a460a0a227dbdb39db766f1159eaea929c9247c7118a0b5bfa58 | 16,386 | null | """
Official Submission for One Layer Deeper Challenge
Tailored Exploration: var1_attention_force
"""
from __future__ import annotations
import math
from typing import Optional, Tuple
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import Optimizer
try:
from benchmark import Su... | {
"id": "44865a0b-344c-486c-96fc-fc8f81e1a8ef",
"created_at": "2026-08-15 21:28:14.397626+00:00",
"db_md5": "3543616e790e5c9fc53f638e52830df3",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "3954f6c8-f276-4760-8d7d-60d7dbfa5757",
"tier": "easy",
"dataset_id": "e1",
"status": "failed",
... | null
|
448a82bc-4086-4bef-812d-9a3a714ab76f | easy | sapient-sapiens | 2026-08-15 20:50:22.027473+00:00 | succeeded | ae1d9d9680e9968770f728476fb2063da8af83fae2dc6598fdb449964ecd9f7d | 21,402 | null | """T-only adaptive chunk routing over a learned shared operator.
The controller sees only the categorical tokens inside the T prompt field and
is trained solely by final-answer loss. It never parses T numerically and T is
not used as an intermediate supervision target. Inputs use learned additive
token, semantic-gro... | {
"id": "448a82bc-4086-4bef-812d-9a3a714ab76f",
"created_at": "2026-08-15 20:50:22.027473+00:00",
"db_md5": "fa40672ef26b85fa1966250eaa2856d4",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "30b364cc-6b08-4436-9589-1a7217d4a2ba",
"tier": "easy",
"dataset_id": "e6",
"status": "su... | {
"score": {
"mean_loss": 4.9609387639450695,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.11013514407583186
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.053988933563232,
"example_count": 60,
... |
448cece7-4bae-44a0-afe2-e2935917d6a3 | easy | velocizapkar | 2026-08-06 22:22:43.908428+00:00 | succeeded | 1f2da1a0e16f9bc661bd2aa5f528a2f8823da064dc6597d2700ba79ab8cba67e | 55,678 | 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": "448cece7-4bae-44a0-afe2-e2935917d6a3",
"created_at": "2026-08-06 22:22:43.908428+00:00",
"db_md5": "bc5d44d14e7651403e28fecc8b4ae70a",
"submitter": "Aakanksh Zarapkar",
"github_login": "velocizapkar",
"run_id": "1af6599e-4c31-45a6-951c-8e364eead6d9",
"tier": "easy",
"dataset_id": "e3",
"statu... | {
"score": {
"mean_loss": 3.1900631189346313,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.01749999914318323
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.169229507446289,
"example_count": 800,... |
44af0bf7-4b99-478e-ac51-96b098f542d0 | easy | Avyukth | 2026-08-19 14:42:44.007840+00:00 | succeeded | 9f3975b8b6841611ed099a6a89ea5bb8fa24aa982dfebd30bdf1db7bd5f50334 | 28,456 | null | """v16 — general-N T-dependent torsion-orbit augmentation (T-503 generalized).
T-503 (hardcoded N=323 kernels) scored 35.00% on hosted Easy e1 — the
strongest generalization signal of the campaign. This version computes the
same 2-power torsion kernels for ANY prompt-parsed modulus at runtime:
factor N (Miller-Rabin ... | {
"id": "44af0bf7-4b99-478e-ac51-96b098f542d0",
"created_at": "2026-08-19 14:42:44.007840+00:00",
"db_md5": "3c4d2ea1ef3e6d87313fb9ff42e071d0",
"submitter": "Subhrajit Makur",
"github_login": "Avyukth",
"run_id": "c2d64110-2e62-4009-8634-89ef64b11daa",
"tier": "easy",
"dataset_id": "e1",
"status": "su... | {
"score": {
"mean_loss": 1.9615800380706787,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.03333333320915699
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.0251150131225586,
"example_count": 100... |
44b080d7-bcd3-4809-8fd7-1644a6938b55 | easy | yunjiangster | 2026-08-26 20:46:38.580431+00:00 | succeeded | bad6eb177365b0bd12bc56f861e1c08f61c6ed74508f852b7ddd6dab65f54b1b | 26,644 | null | """Quantized Markov Digit Machine for the One Layer Deeper benchmark.
The model parses the public prompt format into fixed-width decimal tapes, applies
one learned, tied transition once per requested time step, quantizes the residue
back to categorical digits after every step, and right-aligns the resulting
state with... | {
"id": "44b080d7-bcd3-4809-8fd7-1644a6938b55",
"created_at": "2026-08-26 20:46:38.580431+00:00",
"db_md5": "cacae4e62dd77d56e5f37ad9e93e57a2",
"submitter": "Yunjiang Jiang",
"github_login": "yunjiangster",
"run_id": "5e8fc594-da39-44e0-91d8-232d429e11e0",
"tier": "easy",
"dataset_id": "e6",
"status":... | {
"score": {
"mean_loss": 0.8710374534130096,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.7828829288482666
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.493644654750824,
"example_count": 60,
... |
44b89b40-0b07-4eb4-89bc-cb0c11009830 | easy | viridale | 2026-08-16 03:57:27.672646+00:00 | succeeded | 19d07d9030dcba5927ae45e13f087d19935191a93e686c7680777deaeed611fd | 18,061 | null | """Integer-polynomial boundary quotient with clip-invariant acquisition.
hypothesis: quotienting the T1 coordinate fiber to one autonomous polynomial neuron
and projecting its learned coefficients to integers makes the acquired boundary
compose exactly and transfer across modulus identities.
axis: arch
target:... | {
"id": "44b89b40-0b07-4eb4-89bc-cb0c11009830",
"created_at": "2026-08-16 03:57:27.672646+00:00",
"db_md5": "aa64c42e392b1cd8e58ee153383a4ab4",
"submitter": "priormancer",
"github_login": "viridale",
"run_id": "69485a35-972b-48dc-ba40-f62230f9bfc2",
"tier": "easy",
"dataset_id": "e5",
"status": "succe... | {
"score": {
"mean_loss": 0.15612033125944436,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.15417981147766113,
"example_count": 600,
"e... |
44c75b18-04e6-44a7-8f11-3f8551c36ce9 | easy | chad-atexpedient | 2026-08-27 02:53:11.471904+00:00 | failed | f25893e00d02b93194f8be204a94d913ce2f3f23637740cbdd449331eefcd66d | 19,392 | null | """X132: direct paired-successor supervision.
For supplied same-N chains where A.output equals B.input, decode one shadow
successor from A's terminal state and train it directly on B's supplied output.
This replaces weak latent-only alignment with the exact evaluator-provided
behavioral target. No arithmetic or genera... | {
"id": "44c75b18-04e6-44a7-8f11-3f8551c36ce9",
"created_at": "2026-08-27 02:53:11.471904+00:00",
"db_md5": "312d4c2c868e204dc92667d4e6d453d1",
"submitter": "chad-atexpedient",
"github_login": "chad-atexpedient",
"run_id": "5baed1e3-9b2e-4bdb-8868-7e2a529d368b",
"tier": "easy",
"dataset_id": "e5",
"st... | null
|
44d3386b-9b59-4790-8a0f-287df4197002 | easy | yashkant | 2026-08-31 19:31:45.285613+00:00 | succeeded | d1bee0a7852465f99c6e3402c20a48b2154d0506157ef7f9d9428bb6a542c0d1 | 108,999 | null | """One Layer Deeper submission -- a learned register machine.
Architecture (all outputs produced by learned parameters; no task arithmetic on the answering path):
a bank of R=6 digit registers (W=16 positions each) carried as DISCRETE tokens between micro-steps: five
data registers and one COUNTDOWN register holding t... | {
"id": "44d3386b-9b59-4790-8a0f-287df4197002",
"created_at": "2026-08-31 19:31:45.285613+00:00",
"db_md5": "c7d3c9c69ef15534c69546a26870813a",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "5cb994cc-5718-48bc-bbfb-dc68322631ea",
"tier": "easy",
"dataset_id": "e1",
"status": "succeed... | {
"score": {
"mean_loss": 5.499281883239746,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.013333333190530539
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 5.745937347412109,
"example_count": 100,... |
44d4dd32-9634-4c52-bfe7-8588441a3511 | easy | erdavis0 | 2026-08-27 01:41:51.182258+00:00 | succeeded | f696b4a0dbdab40d944b3bb47ef7b8e716ff6e0849099f9cf030611313221c72 | 31,312 | null | """Typed-prefix reversible transition B3: sequence mixture plus balance.
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": "44d4dd32-9634-4c52-bfe7-8588441a3511",
"created_at": "2026-08-27 01:41:51.182258+00:00",
"db_md5": "aa01dfb95948365c687b75f02f03cd86",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "b9c8664b-70df-475c-9b4b-6f0bab0c53f5",
"tier": "easy",
"dataset_id": "e3",
"status": "succeeded",... | {
"score": {
"mean_loss": 5.336242794668488,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.011875000018626451
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 5.431889469201293,
"example_count": 800,... |
44d6d7e0-682a-4b75-89ea-b99da9e9d343 | easy | richardcepka | 2026-08-13 23:00:29.591583+00:00 | succeeded | f373fb578c426957d11b858822bf59f98a6d7746d245c9a5a7dfb1e2bfaa93c8 | 16,006 | 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": "44d6d7e0-682a-4b75-89ea-b99da9e9d343",
"created_at": "2026-08-13 23:00:29.591583+00:00",
"db_md5": "1635c76300242914a43d7aef8e559967",
"submitter": "Richard Cepka",
"github_login": "richardcepka",
"run_id": "12b018d9-3601-4c11-b32a-86c4f4de44b3",
"tier": "easy",
"dataset_id": "e5",
"status": ... | {
"score": {
"mean_loss": 2.252334571065935,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004166666699262957
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2587125301361084,
"example_count": 600... |
44e492f7-2024-4d24-a049-7a5331019200 | easy | chad-atexpedient | 2026-08-09 21:43:20.280705+00:00 | succeeded | b97e39010bcefe6d24f8bd2bb5495e9d5e53c5327f82f965c46c0efd139c88e9 | 10,027 | null | """BL5A: BL1 post-400 LR 1e-4 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": "44e492f7-2024-4d24-a049-7a5331019200",
"created_at": "2026-08-09 21:43:20.280705+00:00",
"db_md5": "2dcbd32afea71272821b5401c12cfb5e",
"submitter": "chad-atexpedient",
"github_login": "chad-atexpedient",
"run_id": "6d670b9d-4e5b-467a-bf2b-168145dff537",
"tier": "easy",
"dataset_id": "e5",
"st... | {
"score": {
"mean_loss": 3.1602183297507964,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004583333333333333
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.134746252154021,
"example_count": 600... |
44eb661f-2564-4ee6-a725-9cf3a0c7e6b3 | easy | DDanlov | 2026-08-15 21:29:07.194152+00:00 | succeeded | 13975e32744f29f0b4d27f9f311d4fac2d22f5038acddf53e6883962bb8e8a98 | 15,214 | null | """
Official Submission for One Layer Deeper Challenge
Tailored Exploration: var4_log_spectrum
"""
from __future__ import annotations
import math
from typing import Optional, Tuple
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import Optimizer
try:
from benchmark import Submi... | {
"id": "44eb661f-2564-4ee6-a725-9cf3a0c7e6b3",
"created_at": "2026-08-15 21:29:07.194152+00:00",
"db_md5": "12a33577ced010453f4fad7f5cadac9a",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "326094f1-20e6-4b7e-880c-71101f78c034",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.4624682975731966,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004583333333333333
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.4757402799856503,
"example_count": 60... |
44f8644e-a0cf-40f5-8b05-96e4fb6600a6 | easy | sapient-sapiens | 2026-08-18 19:40:07.323845+00:00 | succeeded | 409e8baf080a4dccc21e0e196f290c11712557a961cf2d5a4032dd38516d098c | 31,707 | 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": "44f8644e-a0cf-40f5-8b05-96e4fb6600a6",
"created_at": "2026-08-18 19:40:07.323845+00:00",
"db_md5": "4e9d07298f3be899224809f5a201e005",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "303efb98-5f9f-4574-a169-4cacfb40b8e5",
"tier": "easy",
"dataset_id": "e5",
"status": "su... | {
"score": {
"mean_loss": 2.6564723272891686,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.00791666670391957
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.6185058724185275,
"example_count": 600... |
44fabafe-a938-42e0-8408-3b1836443db1 | easy | sapient-sapiens | 2026-08-17 18:47:26.230006+00:00 | succeeded | ea6191fd6b239ad6edede881c4f1f194680093b426743be7e7cd90a13bf22f84 | 28,671 | null | """v40 lineage: T-free operator, N memory, PonderNet mix, no T in the loss.
Field-local encoder, T-free core, countable halt, answer-slot readout.
N is seeded and restored every step. T is never parsed in forward or loss.
"""
from __future__ import annotations
import math
import time
import torch
import torch.nn.fu... | {
"id": "44fabafe-a938-42e0-8408-3b1836443db1",
"created_at": "2026-08-17 18:47:26.230006+00:00",
"db_md5": "c77d92ed6bf5d40fad9666ac948246dc",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "6d3f4acd-0dd6-46a9-a436-2a9f34a3e9f8",
"tier": "easy",
"dataset_id": "e9",
"status": "su... | {
"score": {
"mean_loss": 3.5110743108864693,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0577777790526549
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.5497591495513916,
"example_count": 90,
... |
4500fe3b-2ca0-4972-abde-5ff32f630249 | easy | shreyash-chonkie | 2026-08-26 17:26:32.917389+00:00 | failed | 2468677e32c1e002c6135f5e8413570bf43c895026bde9f29a87b636dea6f91e | 17,833 | null | """Online hard controller over rank-one associative memory updates."""
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,
)
P... | {
"id": "4500fe3b-2ca0-4972-abde-5ff32f630249",
"created_at": "2026-08-26 17:26:32.917389+00:00",
"db_md5": "7509b2b3f470bbd5cdf2cc2a5c344d64",
"submitter": "Shreyash",
"github_login": "shreyash-chonkie",
"run_id": "12329b54-b2ee-4109-b71f-c430299b60a5",
"tier": "easy",
"dataset_id": "e1",
"status": "... | null
|
450c3469-b01c-4cd6-b179-d7924b68b267 | easy | isaac0804 | 2026-08-05 08:58:20.641455+00:00 | succeeded | f836b272089910a4ba0c4460e0669db3ddc6c9b9b00a877b35732b896ca1e908 | 5,186 | null | """Vanilla (non-recurrent, non-weight-tied) deep Transformer -- a control to
isolate "is a single attention+MLP pass too shallow to even use T" from the
looped architecture's own known failure modes.
`baseline_adamw` (this project's original Step-0 baseline) has exactly one
block: one attention+MLP pass has to read N,... | {
"id": "450c3469-b01c-4cd6-b179-d7924b68b267",
"created_at": "2026-08-05 08:58:20.641455+00:00",
"db_md5": "6a4a79008638fac84c6972fac862e86e",
"submitter": "Isaac Yong",
"github_login": "isaac0804",
"run_id": "32a9cefa-fd5a-4121-9a36-f35bac688264",
"tier": "easy",
"dataset_id": "e3",
"status": "succe... | {
"score": {
"mean_loss": 11.909240551082641,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0056250000046566126
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 12.279618737875136,
"example_count": 8... |
450fd554-8a1c-4187-83b9-2d1b61ab5a74 | easy | chuk-uzowihe | 2026-08-07 15:45:12.008576+00:00 | succeeded | 80aeb3f2ca2fdf3d944a51067f7a970866d7ae671fe17cc36887306e9e59058c | 27,974 | null | """One Layer Deeper submission: FPRM over a compressed (N, x, T) lattice.
v6 = v2's architecture (the only version that has generalized on m1: 15.18%
test, 11.6% at unseen T=32) rebuilt on the v5 chassis, with its one fatal
flaw fixed: input numbers are no longer parsed into int64 (silent garbage
past 19 digits) — ins... | {
"id": "450fd554-8a1c-4187-83b9-2d1b61ab5a74",
"created_at": "2026-08-07 15:45:12.008576+00:00",
"db_md5": "b0399a8aabf51c3148515bba7670ce81",
"submitter": "Chuk Uzowihe",
"github_login": "chuk-uzowihe",
"run_id": "f1a7f581-b31e-478c-b11c-9854d8df7315",
"tier": "easy",
"dataset_id": "e1",
"status": "... | {
"score": {
"mean_loss": 39.88730239868164,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0566666666418314
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 43.663326263427734,
"example_count": 100,
... |
4512e925-d6b2-4f73-a983-bffac0d8a9ed | easy | sapient-sapiens | 2026-08-30 21:55:05.167346+00:00 | succeeded | b626fbb3f3cc3f145d2b38f31441f07e426bf9c85346f3d7aed13b9de17f4d97 | 17,466 | null | """Eligible E5 strict-routed direct Horner with two S transitions per A commit.
The S-compute block reads immutable N/X and advances latent S twice. The
independent A-commit block is exposed only as commit_a(A, S): it receives
freshly zeroed N/X lanes, so current operands reach A only through S (or the
recurrent A st... | {
"id": "4512e925-d6b2-4f73-a983-bffac0d8a9ed",
"created_at": "2026-08-30 21:55:05.167346+00:00",
"db_md5": "3ee089cc961c173915c2155bc292bca1",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "253b3a01-9141-4a40-934b-4164e24fdc64",
"tier": "easy",
"dataset_id": "e5",
"status": "su... | {
"score": {
"mean_loss": 2.1625204868047296,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.008333333333333333
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1640481157687748,
"example_count": 60... |
451539c0-30f6-49f2-8146-3d7f5dbe7059 | easy | viridale | 2026-08-08 10:25:58.082541+00:00 | succeeded | e6234d797e71ca1891602e5e787af548cb0660951936e7986da7d75761e95bb9 | 12,155 | null | """Smooth structural curriculum for the learned categorical Neural GPU.
hypothesis: the v321 cell improved current E1 to 11.33% but learns all numeral lengths
and task depths simultaneously. Classic algorithmic Neural-GPU training benefits
from short-to-long curricula. Smoothly emphasize short input numerals... | {
"id": "451539c0-30f6-49f2-8146-3d7f5dbe7059",
"created_at": "2026-08-08 10:25:58.082541+00:00",
"db_md5": "dafacabc008ea8e59f5c111041a5afdd",
"submitter": "priormancer",
"github_login": "viridale",
"run_id": "6dfc549b-a5e8-478f-8efb-0c5c6e038430",
"tier": "easy",
"dataset_id": "e1",
"status": "succe... | {
"score": {
"mean_loss": 2.020073413848877,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.03499999828636646
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.8664183616638184,
"example_count": 100,... |
45177a8c-9cbb-40d4-8aeb-5ec9c9be6f5c | easy | DDanlov | 2026-08-28 18:33:01.389952+00:00 | succeeded | 92bc3cb461eef96472b3b2b578a40b2cfd249c58af7ac9066c2a3dc7c4d1ea25 | 15,213 | null | """
Model A Variant: Baseline Arch1 (K=1, continuous z0 + zt injection, DKS, RohanShampoo)
Config: D=256, H=4, d_ff=1024, T_rec=128, lr=0.0035
"""
import math
import torch
import torch.nn as nn
import torch.nn.functional as F
from dataclasses import dataclass
from typing import Optional, Tuple, Dict, Any, List, Union
... | {
"id": "45177a8c-9cbb-40d4-8aeb-5ec9c9be6f5c",
"created_at": "2026-08-28 18:33:01.389952+00:00",
"db_md5": "8ed106a07037a19772b9c45c898bd42a",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "952a203c-cd02-48d8-87c4-1016f79ba669",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.307869147636658,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.00875
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.314715328906149,
"example_count": 600,
"e... |
452431bb-f839-470a-b6cb-dce2266a0fc1 | easy | reinhack | 2026-08-28 02:48:34.781510+00:00 | succeeded | 5d61c635ab87a29dc23dfc1d1e59cb6074c0816f3ea86fc4ad0560b25faf0e36 | 28,843 | null | """One Layer Deeper submission: Recurrent Digit-State Squarer (RDS).
The task family is repeated modular squaring: given (N, x, T) as decimal
digit tokens, output x^(2^T) mod N as tail-aligned decimal digits.
Design
------
The model is a learned squaring step iterated T times over an explicit
decimal digit-string sta... | {
"id": "452431bb-f839-470a-b6cb-dce2266a0fc1",
"created_at": "2026-08-28 02:48:34.781510+00:00",
"db_md5": "7b51482950dc2a571a1eed92f162ca48",
"submitter": "reinhack",
"github_login": "reinhack",
"run_id": "f9ff89ac-1743-47c3-9e28-52a350da5413",
"tier": "easy",
"dataset_id": "e1",
"status": "succeede... | {
"score": {
"mean_loss": 2.1521076560020447,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.513333335518837
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.772034764289856,
"example_count": 100,
... |
4525c381-e88e-4892-ab18-34455e08eb84 | easy | jordanrubin | 2026-08-08 10:25:00.565331+00:00 | succeeded | 19e434c7e9fc9d9d601292a0f5c0a4083813bc7790caf87c0582fc1bbcc2cad8 | 16,909 | null | """Autonomous digit-state recurrent Transformer for repeated modular squaring.
The model is deliberately organized around one learned transition:
p_0 = right_aligned_decimal_digits(x)
p_{k+1} = F_theta(p_k, decimal_digits(N))
The same two-block Transformer cell is applied exactly T times. T is used only
as ... | {
"id": "4525c381-e88e-4892-ab18-34455e08eb84",
"created_at": "2026-08-08 10:25:00.565331+00:00",
"db_md5": "671c15c2db39c821999bae96a7defa51",
"submitter": "Jordan Rubin",
"github_login": "jordanrubin",
"run_id": "67cf787e-2574-4318-b521-885f89977f82",
"tier": "easy",
"dataset_id": "e5",
"status": "s... | {
"score": {
"mean_loss": 4.30275765336851,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.009166666384165485
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.323156355608739,
"example_count": 600,
... |
452905a8-8fb5-40aa-b98d-531260d5c0c1 | easy | Ranoobaba | 2026-08-19 13:21:17.050667+00:00 | succeeded | b71314738962d767d7c6eeb1c9d67deabd95a17ff4fcfd2b69ef853953d51498 | 6,601 | null | """Looped transformer, iteration count conditioned on T read from the prompt.
Core idea: learn one latent "squaring step" operator and apply it T times.
Depth generalization comes from the loop structure itself; the model never
needs more parameters for deeper problems, only more iterations.
"""
from __future__ impor... | {
"id": "452905a8-8fb5-40aa-b98d-531260d5c0c1",
"created_at": "2026-08-19 13:21:17.050667+00:00",
"db_md5": "462bd1aee09ff4498696aeaa01594429",
"submitter": "Syed Rayyan Ali",
"github_login": "Ranoobaba",
"run_id": "96abbe76-4f5b-47c2-affd-0328e4d6dfb0",
"tier": "easy",
"dataset_id": "e3",
"status": "... | {
"score": {
"mean_loss": 6.275145166979218,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.01125
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 6.3411995579226454,
"example_count": 800,
"... |
452b9279-71ff-4978-8822-f6533ac87b1d | easy | blake-camp-surge | 2026-08-04 03:09:46.232431+00:00 | succeeded | be6c847a5c839dbce932da69c0bb290a56367ed5de84f6ee01dbb710691c9947 | 12,490 | null | """Vector DANs communicating through recurrent linear-attention memory."""
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,
a... | {
"id": "452b9279-71ff-4978-8822-f6533ac87b1d",
"created_at": "2026-08-04 03:09:46.232431+00:00",
"db_md5": "d8bd054d161e87679e9bc0d3065ceca3",
"submitter": "Blake Camp",
"github_login": "blake-camp-surge",
"run_id": "4b669b14-12ed-4ea4-bc47-9db46ca6243d",
"tier": "easy",
"dataset_id": "e3",
"status":... | {
"score": {
"mean_loss": 2.1395671484952716,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.016875000093132256
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1522411185751253,
"example_count": 80... |
453641fa-74da-482b-8ffa-6d63abe6a100 | easy | AFeinfield | 2026-08-28 04:02:12.921046+00:00 | succeeded | 75e1fd8d7df9293d7a9a7a7d7ffbf59a4d7a1d2e48d93476b0a008ee97ce6b39 | 11,546 | null | """184: 182 recipe arm gf_lamb=1.0 wd=1.0.
182 notes: SPELL-HEAD AUX (head learns to spell every residue seen
as an input, from the prompt's own digit tokens — pure re-encoding).
Targets the residual ~15%: held-out answers whose residues never appear
as training answers were unspellable however well F routed.
Local p... | {
"id": "453641fa-74da-482b-8ffa-6d63abe6a100",
"created_at": "2026-08-28 04:02:12.921046+00:00",
"db_md5": "30c738e1e5b776c8e74adc8c55e864b7",
"submitter": "AFeinfield",
"github_login": "AFeinfield",
"run_id": "4b861fb3-4647-44a5-8d0b-e2190a5d7052",
"tier": "easy",
"dataset_id": "e5",
"status": "succ... | {
"score": {
"mean_loss": 2.8220948873693747,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.1004166666790843
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.7772055677234326,
"example_count": 600,... |
4539531f-68ab-437b-bd35-814b3928775c | easy | dlyr3 | 2026-08-23 19:34:59.369932+00:00 | failed | f94b635f6dd2b70c17e8a3d599ceaea226881d1a6201c012187cb986cfacb760 | 7,386 | null | """The M-layer of arXiv:2008.03936 applied to the repeated-squaring task.
M = B + T(U x) is built from the pooled prompt, exp(M) is taken with
torch.matrix_exp, and V + S exp(M) is read back out. The paper's activity regularisation is deliberately NOT applied: measured on e1,
any nonzero lambda drives ||exp(M)||_F... | {
"id": "4539531f-68ab-437b-bd35-814b3928775c",
"created_at": "2026-08-23 19:34:59.369932+00:00",
"db_md5": "3f569f9edd87fb3629c5fd5274b314cf",
"submitter": "dlyr3",
"github_login": "dlyr3",
"run_id": "9e2211fd-fda6-4ab7-ba48-947b4e939705",
"tier": "easy",
"dataset_id": "e1",
"status": "failed",
"st... | null
|
453b6aa8-773c-4f66-b308-5920a684ccbc | easy | DDanlov | 2026-08-23 22:25:33.101431+00:00 | failed | 33d965c65d03c3cd63d04308d59765f531d38f13f6073c46343d4adfc93fa7cc | 32,828 | null | """
Unified Hybrid Masked Deliberation Model (40.4M Parameters) - Pure Skipless, No Cross-Attention
Architecture:
- Scale: dim=1536, num_heads=24, d_ff=6144 (Exactly 40,437,888 parameters)
- Hybrid Prefix-Bidirectional + Causal Target Attention Mask
- Token Layout: [Prompt (L_p), 4 Pause Registers, Target Tokens (L_tgt... | {
"id": "453b6aa8-773c-4f66-b308-5920a684ccbc",
"created_at": "2026-08-23 22:25:33.101431+00:00",
"db_md5": "b1b030dd1a2dcd4533506d0256ff5ceb",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "ab61aac5-9cdd-4d6c-851c-03e161d1f3c2",
"tier": "easy",
"dataset_id": "e5",
"status": "failed",
... | null
|
453d1e5f-db1c-4b75-999b-831c755f1f2f | easy | Dandandan | 2026-08-04 11:59:11.678936+00:00 | succeeded | 0f2ebcfbb4b20fee9fa434d360607c5d0f946744ee69e3740d445c6c973f38de | 12,180 | null | """Learned spectral-recurrence submission for One Layer Deeper.
The recurrence is not selected in code. A randomly initialized spectral bank
learns the one-step map from examples and the same learned map is rolled out for
larger T. Decimal parsing and rendering only define the task's token interface.
"""
from __fut... | {
"id": "453d1e5f-db1c-4b75-999b-831c755f1f2f",
"created_at": "2026-08-04 11:59:11.678936+00:00",
"db_md5": "4c75f0a04425a4ae98ed84c5dff46d7c",
"submitter": "Daniël Heres",
"github_login": "Dandandan",
"run_id": "e5ffdf2e-1f59-4470-9e2a-464298fa7aa2",
"tier": "easy",
"dataset_id": "e1",
"status": "suc... | {
"score": {
"mean_loss": 1.1271511084487429e-05,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.1280336366326082e-05,
"example_count": 100,
... |
454edb29-73a9-400a-89a6-b664d0e0f17a | easy | khushidahi | 2026-08-20 03:43:37.100934+00:00 | succeeded | bb5b2081f03d9d4ccb4746c83247d1c25d5ff98c2928870d3d33ff0077e19f46 | 32,116 | null | """R295: tied causal two-stage categorical quadratic functional root.
This candidate is deliberately diagnostic. It trains on evaluator-owned final
labels at several depths so that one shared transition must compose. The model
never converts N or X to native numbers and never implements multiplication or
modular red... | {
"id": "454edb29-73a9-400a-89a6-b664d0e0f17a",
"created_at": "2026-08-20 03:43:37.100934+00:00",
"db_md5": "ae879e8f3646417e7290d6f087e8fed8",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "d0864a9f-3e59-42df-877e-180a6532cf70",
"tier": "easy",
"dataset_id": "e3",
"status": "succ... | {
"score": {
"mean_loss": 2.269275233561568,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0068750000093132264
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.253095573771135,
"example_count": 800... |
45509933-ab4e-4afd-b1f0-b8069a71c49f | easy | jayyala | 2026-08-16 22:25:36.920406+00:00 | succeeded | b5acf766cf76f2016093f66fe292fb0198f4803d7ccb9de5f3a468aec4dbb9ce | 13,296 | null | """A scalable recurrent machine with canonical categorical byte states.
Each integer is represented by a short little-endian tape of 256-way slots.
One tied bidirectional mixing cell learns the unknown state transition from the
current tape and the prompt context, then its categorical output is fed back
for exactly T ... | {
"id": "45509933-ab4e-4afd-b1f0-b8069a71c49f",
"created_at": "2026-08-16 22:25:36.920406+00:00",
"db_md5": "10dc525bee82f997424fae858ca3f88b",
"submitter": "jayyala",
"github_login": "jayyala",
"run_id": "e5485d03-3313-468c-b462-cb37ccaa0597",
"tier": "easy",
"dataset_id": "e6",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.246575355529785,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.9358108341693878
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.0,
"example_count": 60,
"exact... |
4551be51-d071-49e3-8247-ee2bca502f85 | easy | viridale | 2026-08-03 21:34:59.550780+00:00 | succeeded | 4cda0ca962056cf33a776d62d228d1dc0ee5863efb6b881a441e94825634814b | 144,966 | null | """CUDA-wide mismatch-triggered broad evidence scan.
hypothesis: H100 fallback latency is dominated by Python/kernel launch overhead;
a CUDA-only 4194304-atom chunk uses measured memory headroom to reduce it.
axis: arch
target: revised h1 Max T at least1 with faster unsupported-route commitment.
expected_delta: pr... | {
"id": "4551be51-d071-49e3-8247-ee2bca502f85",
"created_at": "2026-08-03 21:34:59.550780+00:00",
"db_md5": "990f3a734d7da60a1d7b7352bea851f2",
"submitter": "priormancer",
"github_login": "viridale",
"run_id": "bc00c95b-f64f-43a4-806b-55a2b7d468aa",
"tier": "easy",
"dataset_id": "e2",
"status": "succe... | {
"score": {
"mean_loss": 0.0030145463443549637,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.0030145463206492683,
"example_count": 300,
... |
45577916-4ddc-44b5-aa8e-74c77cfcfb31 | easy | oupadhyay | 2026-08-15 22:11:58.190680+00:00 | succeeded | a6cdb33fbb13eb704cb5a36d1a713e4eb7f8b4b80843ebd132b9c60062378f24 | 3,994 | null | """Conditional optimizer phase 2: lr10_wd3."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import nn
from benchmark import ModelSpec,OptimizerSpec,OptimizerBundle,Submission,assert_model_state
W,D,H=4,64,32
class C:
def __init__(s,vocab_size,max_seq_len):s.vocab_s... | {
"id": "45577916-4ddc-44b5-aa8e-74c77cfcfb31",
"created_at": "2026-08-15 22:11:58.190680+00:00",
"db_md5": "9b7e1b3faf681688992939eae0de717d",
"submitter": "Ojasw Upadhyay",
"github_login": "oupadhyay",
"run_id": "21a30b2d-63a9-458d-82cb-5b16324f72cf",
"tier": "easy",
"dataset_id": "e5",
"status": "s... | {
"score": {
"mean_loss": 2.4095004658709778,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0025
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.3144280910491943,
"example_count": 600,
"... |
455a29f3-c3ce-4a2d-8bbb-066c7819a422 | easy | KaustubhKumar05 | 2026-08-25 15:33:55.234451+00:00 | succeeded | cede8f877bf90dc9d38af3879dc3680a73b6693631c3476085830cfc30a955d0 | 16,027 | null | """final_v1: culmination candidate. v8 (N-FiLM) + Muon momentum 0.5 + tail-EMA.
A small MLP embeds the modulus digits (mean-pooled over the N span) and
emits per-channel scale/shift applied to the cell input and a bias on the
update gate, every iteration. Hypernetwork-lite: the executor is modulated
by N rather than h... | {
"id": "455a29f3-c3ce-4a2d-8bbb-066c7819a422",
"created_at": "2026-08-25 15:33:55.234451+00:00",
"db_md5": "4c081115900db9bcea126ae905aeee44",
"submitter": "koz",
"github_login": "KaustubhKumar05",
"run_id": "e717bb97-88ff-4253-98c5-57c84e830ff2",
"tier": "easy",
"dataset_id": "e3",
"status": "succee... | {
"score": {
"mean_loss": 6.7707328014824935,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.01625
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 7.270838411304697,
"example_count": 800,
"... |
455dd9a4-1e59-4bc5-90d8-2892e999cbc9 | easy | sapient-sapiens | 2026-08-18 10:55:54.505315+00:00 | succeeded | 02f48d2937dc39dc1bdb9e402d573b1111129ebf16e6174aa3e340f31563b92e | 26,973 | null | """v44 adaptive baseline: one encoder, four-layer T-free core, bounded halt.
Locked routing: field-local one-layer encoder; recurrent core sees only N/X;
T is pooled solely for the controller; decode once on a halt-weighted state
mix. Stopping support is 8 depths, not 64. T is never parsed in forward or
the loss.
"""
... | {
"id": "455dd9a4-1e59-4bc5-90d8-2892e999cbc9",
"created_at": "2026-08-18 10:55:54.505315+00:00",
"db_md5": "b5c285cfb1b1680da8c473fff9b0b2cd",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "d787b813-8d5c-4899-8f98-498da607e42b",
"tier": "easy",
"dataset_id": "e9",
"status": "su... | {
"score": {
"mean_loss": 1.894828896678665,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.034444444750746095
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.8853759765625,
"example_count": 90,
... |
456052e1-21aa-401c-9c7a-9781599d030b | easy | viridale | 2026-08-12 05:42:08.107132+00:00 | succeeded | d0d367e50d7fd3f44168d68b20bccfa8d8f63b62ee523e8dd68d256c32a1dbdc | 62,682 | null | """Learned low-bit control plane over the exact recurrent integer carrier.
hypothesis: v516's continuous integer law can express polynomial, phase, event,
wrapper, and two-register edits, but every transition and clock remains affine in
prompt-level control and every reduction is fixed to N. Add a generic lear... | {
"id": "456052e1-21aa-401c-9c7a-9781599d030b",
"created_at": "2026-08-12 05:42:08.107132+00:00",
"db_md5": "f41644ac5352d7365c8db631e7b7bbf5",
"submitter": "priormancer",
"github_login": "viridale",
"run_id": "3451372e-d420-49d8-af50-8eb72fbb6fa7",
"tier": "easy",
"dataset_id": "e1",
"status": "succe... | {
"score": {
"mean_loss": 0.15925879776477814,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 0.15121066570281982,
"example_count": 100,
"e... |
45609093-a466-4130-94ba-52b7d6de58ea | easy | mertunsall | 2026-08-06 19:51:51.003686+00:00 | succeeded | ae9c0e8e480c4d94d6903c9231ee3b80827d981a5d40c87b63b6480f67f9648f | 19,382 | null | """Residue-composer baseline with gated recurrent step supervision.
The model parses N, x, and T, builds learned hash/numeric features, predicts a
direct distribution over small residue values, and also runs a tied transition
core from an x/N state for a bounded number of latent steps. The two value
distributions are... | {
"id": "45609093-a466-4130-94ba-52b7d6de58ea",
"created_at": "2026-08-06 19:51:51.003686+00:00",
"db_md5": "4494756f6e8753408b006783df72bf3b",
"submitter": "Mert Unsal",
"github_login": "mertunsall",
"run_id": "9b8d07e1-76fe-4065-a03c-31ac18ba7a94",
"tier": "easy",
"dataset_id": "e5",
"status": "succ... | {
"score": {
"mean_loss": 33.839150815288335,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0320833333581686
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 33.83145456356853,
"example_count": 600,
... |
45680722-0b5e-49e6-a0ce-f9c99d17234e | easy | repst | 2026-08-24 22:33:30.933163+00:00 | succeeded | b7479b002823c0dcfb130b5d446ade083bf6151cb982b96d244ce1e6225c2b85 | 34,364 | null | # R641: cached three-instruction semantic-frontier register machine.
#
# R641 GENERALITY/EXPRESSIVITY DELTA:
# A domain-neutral MDL semantic-closure design now supplies up to three serial
# SSA instructions instead of a bespoke two-step constructor. Equivalent syntax
# is quotiented before learning. Training executes o... | {
"id": "45680722-0b5e-49e6-a0ce-f9c99d17234e",
"created_at": "2026-08-24 22:33:30.933163+00:00",
"db_md5": "3330615b5e9343689f828bc9b387a028",
"submitter": "Ryan Epstein",
"github_login": "repst",
"run_id": "9b9d4193-fc89-4075-8936-cb96bfdc21a1",
"tier": "easy",
"dataset_id": "e1",
"status": "succeed... | {
"score": {
"mean_loss": 2.211726614959222,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.04333333276212215
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.204557418823242,
"example_count": 100,
... |
4569061b-9f47-4621-aae6-c28d5d3659b0 | easy | sapient-sapiens | 2026-08-20 15:33:53.610135+00:00 | succeeded | 0c0a3fe3363bbd71371446fb3add6f96fd2008b94ad7b0f2d923a3152bac4110 | 38,861 | null | """Closed learned spectral interface around a categorical recurrent operator.
The recurrent state is a bank of right-aligned categorical digit slots. Each
application uses a shared, shallow N-to-X conditioner to create learned phase
coordinates, injects paired sin/cos features as a small residual, applies the
ordinar... | {
"id": "4569061b-9f47-4621-aae6-c28d5d3659b0",
"created_at": "2026-08-20 15:33:53.610135+00:00",
"db_md5": "8c774f8d3eaf9e85152ab098a32d6710",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "7e3a507b-7f3e-4805-8328-3e9550c2ced9",
"tier": "easy",
"dataset_id": "e5",
"status": "su... | {
"score": {
"mean_loss": 2.2009148440968547,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.002916666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1976434596450876,
"example_count": 60... |
456bc8d3-b8e1-43cb-b696-f61aa17161f7 | easy | alirezashirvani-jr | 2026-08-31 23:02:53.288592+00:00 | succeeded | ee5676a0db8bd75462ea4cb8aea4139e8755573ebaf4f612203d90ad955ccd39 | 25,614 | null | from __future__ import annotations
import math
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,
)
HP = {
... | {
"id": "456bc8d3-b8e1-43cb-b696-f61aa17161f7",
"created_at": "2026-08-31 23:02:53.288592+00:00",
"db_md5": "cccbc44eebed8e67ac33f86f5cc4c1f7",
"submitter": "alirezashirvani-jr",
"github_login": "alirezashirvani-jr",
"run_id": "7b01fca6-b70c-42ab-8a8f-5453f2373304",
"tier": "easy",
"dataset_id": "e5",
... | {
"score": {
"mean_loss": 2.619612686590835,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0037500000124176345
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.6122901351462566,
"example_count": 60... |
456bd9dd-82af-4db5-9a4c-f7bead23f10e | easy | alirezashirvani-jr | 2026-09-01 00:49:29.779539+00:00 | succeeded | e133e8366d2dfb00a35a6d4af6445d4ed59ff8b55b7b63e65c1d81a4f2a62bc2 | 20,873 | null | from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
assert_model_state,
)
PAD = 0
MODULUS = 2
VALUE = 3
DEPTH = 4
ANSWER = 5
EO... | {
"id": "456bd9dd-82af-4db5-9a4c-f7bead23f10e",
"created_at": "2026-09-01 00:49:29.779539+00:00",
"db_md5": "5791127579d210d47794be9891c5b509",
"submitter": "alirezashirvani-jr",
"github_login": "alirezashirvani-jr",
"run_id": "1d9f8802-ec9d-4387-a85f-b9a51317bdb8",
"tier": "easy",
"dataset_id": "e5",
... | {
"score": {
"mean_loss": 2.777728773687971,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.010416666666666668
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.7863930757804836,
"example_count": 600... |
456c2802-e704-4c0d-aaad-cc1f7b2514a3 | easy | erdavis0 | 2026-08-22 02:41:02.582356+00:00 | succeeded | dd717b6474fda04ebdb5466cb2037b63cbf06831fdcb8eb294b90d300288916b | 5,797 | null | """Soft field factorization with multiplicative field interactions."""
from __future__ import annotations
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
assert_model_state,
)
WIDTH = 64
FIE... | {
"id": "456c2802-e704-4c0d-aaad-cc1f7b2514a3",
"created_at": "2026-08-22 02:41:02.582356+00:00",
"db_md5": "464075811335119d26ad5014b80665e5",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "a3e419ad-aedc-4000-b4b5-cdbfb99a456b",
"tier": "easy",
"dataset_id": "e7",
"status": "succeeded",... | {
"score": {
"mean_loss": 9.571274757385254,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.02103386865928769
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 9.50877571105957,
"example_count": 85,
... |
457c281f-9629-425b-8e2a-822e38eb40be | easy | AFeinfield | 2026-08-22 14:45:12.759917+00:00 | succeeded | 13b18126cac1d39e27cb85c362a5e538b57a3ed7e5fecdb72d0dc4430e26df88 | 19,306 | null | """092: INSTRUMENTED BASE — two measurement fixes, no architecture change.
(1) Step-indexed LR schedule with a fixed screening step budget. The old
schedule decayed on WALL CLOCK, so throughput jitter changed the
LR-per-step trajectory: identical files on e2 gave 1470 vs 1025 steps
and train acc 0.725 vs 0... | {
"id": "457c281f-9629-425b-8e2a-822e38eb40be",
"created_at": "2026-08-22 14:45:12.759917+00:00",
"db_md5": "fbfde8069d90ebb1b53792fca27f8bb0",
"submitter": "AFeinfield",
"github_login": "AFeinfield",
"run_id": "3a5fc991-6e4c-4732-af7c-22e612bb252c",
"tier": "easy",
"dataset_id": "e10",
"status": "suc... | {
"score": {
"mean_loss": 1.9477409720420837,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.00800000037997961
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.940250277519226,
"example_count": 125,... |
457c590e-d3eb-4668-b0be-9acc600b9761 | easy | gauravmishra | 2026-08-09 18:56:47.521944+00:00 | succeeded | 45b62f7d88fd591ffaf01932914131e03c33873f6f71c67a6af9c648d417c7a3 | 6,078 | null | """Compound-scaled Transformer with distinct blocks and 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_st... | {
"id": "457c590e-d3eb-4668-b0be-9acc600b9761",
"created_at": "2026-08-09 18:56:47.521944+00:00",
"db_md5": "1bd7e29f1a5f7c20fa0be72253b35798",
"submitter": "Gaurav Mishra",
"github_login": "gauravmishra",
"run_id": "b93aff56-5a1e-48f0-b074-827a40e66057",
"tier": "easy",
"dataset_id": "e5",
"status": ... | {
"score": {
"mean_loss": 2.3695691350697135,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004583333333333333
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.4529035262462804,
"example_count": 60... |
45862b7c-5e3f-4566-bc8d-ce1424607d22 | easy | satyaki-up | 2026-08-05 16:00:24.633367+00:00 | succeeded | 055e2ccbdcf594a0f8791843b79e8b72b88949f85ee15aa93309656369ea60bf | 3,792 | 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": "45862b7c-5e3f-4566-bc8d-ce1424607d22",
"created_at": "2026-08-05 16:00:24.633367+00:00",
"db_md5": "f9787284fbe35fa0304749b27924c12a",
"submitter": "Satyaki",
"github_login": "satyaki-up",
"run_id": "fe04368d-628d-4997-9fdc-d88385f9f9d3",
"tier": "easy",
"dataset_id": "e1",
"status": "succeed... | {
"score": {
"mean_loss": 2.833213686943054,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.8332138061523438,
"example_count": 100,
"exac... |
45869441-7c70-4ee5-9cf0-f6275d61f235 | easy | DDanlov | 2026-08-17 12:00:12.711837+00:00 | succeeded | 9bae61b73b4f49536bb29366f65c5726ea425ba53d512b6bbea7352bca69596d | 31,553 | null | """
Scalable Recurrent Deliberation Transformer (ScalableDEPHNet)
============================================================
Architecture:
- Pure No-Norm: Single unit RMS normalization strictly isolated to the input token embedding layer.
Zero LayerNorm / RMSNorm inside recurrent deliberation blocks.
- Pure No-Skip... | {
"id": "45869441-7c70-4ee5-9cf0-f6275d61f235",
"created_at": "2026-08-17 12:00:12.711837+00:00",
"db_md5": "4c598e3ad7c2ff150a10839c4ca75dd0",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "0e5fe3b0-98ad-44e5-9d02-3a0d7240ccb6",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.14374216774158,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.06166666472951571
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2824158668518066,
"example_count": 100,
... |
458832e7-715b-4a1b-8b28-0143216c81fe | easy | lzy54 | 2026-08-27 15:35:46.271550+00:00 | succeeded | d333749ac9a864db1a5b0ae2d3bbb87cf98d4142f72ba15ba441b7cdb8ab2951 | 4,673 | null | """Phase 2 C3: one stem plus a three-pass shared residual-recall core."""
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 = 1... | {
"id": "458832e7-715b-4a1b-8b28-0143216c81fe",
"created_at": "2026-08-27 15:35:46.271550+00:00",
"db_md5": "8983639a461d78e8890732e1981d5ef0",
"submitter": "lzy54",
"github_login": "lzy54",
"run_id": "8083390b-c628-484d-b6a4-a6e1952662cf",
"tier": "easy",
"dataset_id": "e6",
"status": "succeeded",
... | {
"score": {
"mean_loss": 2.358476400375366,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.02837838069535792
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2646307945251465,
"example_count": 60,
... |
458bbe53-cdcd-4e5d-b908-e8bebd5ef5fa | easy | yashkant | 2026-08-12 03:49:04.695684+00:00 | succeeded | 71d755e84cf0fd906cde81e70c9e4ecb6d1ac038cf7a789faac0c34f3dd2f466 | 32,160 | null | """Round1787 contrastive set-code centering on the W16/D1 carrier.
This evolutionary graft keeps the carrier, one shared projection per parsed
T, recurrent feedback, optimizer, LR, EMA, and parameter count fixed. It
removes the shared mean from the two learned Dykstra set codes and doubles the
remaining contrast, mak... | {
"id": "458bbe53-cdcd-4e5d-b908-e8bebd5ef5fa",
"created_at": "2026-08-12 03:49:04.695684+00:00",
"db_md5": "1ad98eed6824aa51aae7cb2f930a0157",
"submitter": "Yash Kant",
"github_login": "yashkant",
"run_id": "fac3e165-a02b-4617-adad-f364fb5f60a4",
"tier": "easy",
"dataset_id": "e2",
"status": "succeed... | {
"score": {
"mean_loss": 3.3756139874458313,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.17750000208616257
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.7805352210998535,
"example_count": 300... |
458d6d20-7eaf-466e-b76f-53810ebf80ea | easy | DDanlov | 2026-08-27 21:43:29.986815+00:00 | succeeded | 79c296684f14fba3cb1c061ec4c7e076436d46fe4be5bf64a759411fb6c3da3e | 38,195 | null | import math
import time
from dataclasses import dataclass
from typing import Optional, Tuple, Dict, Any, List, Union
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import Optimizer
try:
from onelayerdeeper.api import Submission, OptimizerBundle, ModelSpec, OptimizerSpec
except ... | {
"id": "458d6d20-7eaf-466e-b76f-53810ebf80ea",
"created_at": "2026-08-27 21:43:29.986815+00:00",
"db_md5": "95d40c989d9cbb32369dae022ac4c4ec",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "f02db5de-d10d-40f5-87c3-4308c857dfe5",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.741710901260376,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004166666651144624
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.743138313293457,
"example_count": 600,... |
458e84b4-c364-4881-bfdf-fcb4f8e9e8c3 | easy | DDanlov | 2026-08-09 00:16:10.376549+00:00 | failed | dd9176bc2344094d159248ae2c347e954728ae0fa63a055969e8ac2788a93848 | 15,978 | null | from __future__ import annotations
import sys
import os
import math
import time
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim.optimizer import Optimizer
VOCAB_SIZE = 17
DIGIT_OFFSET = 7
class RohanShampoo(Optimizer):
"""Self-contained RohanShampoo optimizer with eigenvalue ma... | {
"id": "458e84b4-c364-4881-bfdf-fcb4f8e9e8c3",
"created_at": "2026-08-09 00:16:10.376549+00:00",
"db_md5": "09a1fdf07aa34499a24a33c0ee20373e",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "66dacc73-f859-4c24-8e8d-7e8610340d43",
"tier": "easy",
"dataset_id": "e1",
"status": "failed",
... | null
|
45980956-f389-45be-bfe1-6cfdefff5d14 | easy | DDanlov | 2026-08-09 00:48:01.090920+00:00 | succeeded | 2b39027ad89a3ca20516f816b05c71f03e4804da979089620da593dfe585eed7 | 15,981 | null | from __future__ import annotations
import sys
import os
import math
import time
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim.optimizer import Optimizer
VOCAB_SIZE = 17
DIGIT_OFFSET = 7
class RohanShampoo(Optimizer):
"""Self-contained RohanShampoo optimizer with eigenvalue ma... | {
"id": "45980956-f389-45be-bfe1-6cfdefff5d14",
"created_at": "2026-08-09 00:48:01.090920+00:00",
"db_md5": "5f4617b3d98c01ac509187eab29e4c79",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "ac849d84-79f9-4ec4-8a2d-3063035d9c08",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 83.61446160865842,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0033333333333333335
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 87.22539520263672,
"example_count": 100... |
459b35d0-43c5-4465-8c14-e603245b383b | easy | sapient-sapiens | 2026-08-19 18:19:23.610455+00:00 | succeeded | aa1c7ca6dda776d42b51cca099ef3dc1f6572e673b91d69a9a845b7969843295 | 33,252 | 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": "459b35d0-43c5-4465-8c14-e603245b383b",
"created_at": "2026-08-19 18:19:23.610455+00:00",
"db_md5": "eaadecb09b069ac63ebdde8677fccbdf",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "df05cf93-0147-481b-a20d-96f4e5580c25",
"tier": "easy",
"dataset_id": "e5",
"status": "su... | {
"score": {
"mean_loss": 2.2096841278974058,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.005416666666666667
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2027883502934547,
"example_count": 60... |
459ed329-8979-49ca-a8ec-85230c11143d | easy | erdavis0 | 2026-08-21 11:20:38.589109+00:00 | succeeded | aaf896656a39960006176593f51c42466ac987207d80a5db9c8cd173ee6d2562 | 9,256 | null | """Context-compiled monotone executor with sealed tied recurrence."""
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,
)
WI... | {
"id": "459ed329-8979-49ca-a8ec-85230c11143d",
"created_at": "2026-08-21 11:20:38.589109+00:00",
"db_md5": "92ddd807eb4760c5cd1a4c89a3bb7a3e",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "0cf4ba89-553d-40cb-9e8e-2891a1ddeda7",
"tier": "easy",
"dataset_id": "e4",
"status": "succeeded",... | {
"score": {
"mean_loss": 2.109737614014793,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.005833333358168602
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1139338429911434,
"example_count": 120... |
45a269e5-235c-4055-9772-d021593a603d | easy | pculaf | 2026-08-15 17:58:52.004223+00:00 | succeeded | 97d486677a7c5bd043fbded15d35a12eda161a78be059f6969a1628b10a6f626 | 8,356 | 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": "45a269e5-235c-4055-9772-d021593a603d",
"created_at": "2026-08-15 17:58:52.004223+00:00",
"db_md5": "dc83cf269690eb4c9fe3e04de1922c6c",
"submitter": "Pavle Culafic",
"github_login": "pculaf",
"run_id": "65700b69-2450-4eea-80ee-face9d0cd693",
"tier": "easy",
"dataset_id": "e6",
"status": "succe... | {
"score": {
"mean_loss": 2.0630292892456055,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.006756756920367479
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.095714807510376,
"example_count": 60,... |
45a34ece-33ea-4daf-91c7-7e99ac7b9ab0 | easy | GLauzza | 2026-08-20 21:07:13.121456+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": "45a34ece-33ea-4daf-91c7-7e99ac7b9ab0",
"created_at": "2026-08-20 21:07:13.121456+00:00",
"db_md5": "eaac69fe345f82973fcf392c504a40ff",
"submitter": "Gabriel",
"github_login": "GLauzza",
"run_id": "5bb5eea0-4ab6-41a3-9da4-8227b369900e",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.989136576652527,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.009999999776482582
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.8821585178375244,
"example_count": 100... |
45a78459-7c77-46bd-a4c3-21782c6cc337 | easy | DDanlov | 2026-08-31 21:43:07.487854+00:00 | succeeded | 2ff17ecc36f86cfccd2593567665a9f8b3f9c7266b1b2514f4ef3db2396dd0dc | 32,410 | 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": "45a78459-7c77-46bd-a4c3-21782c6cc337",
"created_at": "2026-08-31 21:43:07.487854+00:00",
"db_md5": "7f8162c2f0e93eb11f6328befd330a70",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "d54eb7d5-3410-486b-9f70-96df12138fb0",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 3.086826363704933,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004583333345750968
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.0863001902541773,
"example_count": 600... |
45a9e007-cfe8-4747-b0d1-114fec653ec1 | easy | liam-gb | 2026-08-11 17:02:04.380744+00:00 | succeeded | a91c83a2dfd13a5dc896db83beb80c57739bf9970ca116d6f8827525f5a45060 | 13,278 | null | """rns_pm_gather_ct: rns_pm_gather + counted halting on a weight-tied
loop.
Changes vs rns_pm_gather: the sigmoid T-gates and per-step embeddings
are DELETED (per-iteration parameters past the trained depth receive no
gradient and are exactly the measured extrapolation gap). One shared
transition is applied hard-count... | {
"id": "45a9e007-cfe8-4747-b0d1-114fec653ec1",
"created_at": "2026-08-11 17:02:04.380744+00:00",
"db_md5": "6905ef43ae1bcceded1feba840f41533",
"submitter": "liam-gb",
"github_login": "liam-gb",
"run_id": "0fb772be-22c3-4737-872e-ad3601bbe8c9",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 4.678142309188843,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.4766666740179062
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 8.329679489135742,
"example_count": 100,
... |
45b7114d-de58-4dec-87e5-3a2cbb612804 | easy | sapient-sapiens | 2026-08-17 02:43:47.092465+00:00 | succeeded | 96d137d90e52e00d4bfe7084329085f05efee0981345708cdf647be0670dae02 | 15,120 | null | """Fixed-4 recurrence candidate: v70 source without adaptive halt.
Parent: act64_e6_w1024_dh4_L6_wd1_v70 (width 1024, d_h=4, L=6, WD=1.0).
Primary variable: adaptive ACT over 1--64 (typically hard-halt 2) versus a
fixed four applications of the shared core. The halt controller is removed.
T remains an ordinary prompt ... | {
"id": "45b7114d-de58-4dec-87e5-3a2cbb612804",
"created_at": "2026-08-17 02:43:47.092465+00:00",
"db_md5": "a360270b1bd62e5c960d2cc497021af4",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "ae0d862c-7b79-4c3b-a05d-ed97ece701bb",
"tier": "easy",
"dataset_id": "e5",
"status": "su... | {
"score": {
"mean_loss": 2.27476426109925,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.005833333358168602
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.2644124776793166,
"example_count": 600,... |
45be8e1b-02ea-40c3-9c98-9b570e018939 | easy | gauravmishra | 2026-08-08 00:12:02.505056+00:00 | succeeded | 9140fef522ec6a9afeb558e69343f534c498c0164ab7648853f9bbf90a742043 | 3,726 | null | """Four-pass Transformer with four independently parameterized blocks."""
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 = 1... | {
"id": "45be8e1b-02ea-40c3-9c98-9b570e018939",
"created_at": "2026-08-08 00:12:02.505056+00:00",
"db_md5": "067dd19f95b3a2b025d36561c727f6b0",
"submitter": "Gaurav Mishra",
"github_login": "gauravmishra",
"run_id": "3ae000cd-d5f6-4342-a941-199a0bd1e7fb",
"tier": "easy",
"dataset_id": "e5",
"status": ... | {
"score": {
"mean_loss": 11.086028110970513,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.006666666728754838
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 11.408707516075784,
"example_count": 60... |
45c06e33-a5ec-4c2a-8271-c32bc6489ce8 | easy | chad-atexpedient | 2026-08-14 15:17:57.895939+00:00 | succeeded | 1deeef70c3b578d25e36d448df9d935b62a0cd6796e31c68a8f1d87eb6ac3c92 | 17,313 | null | """X18: E04 with compositional consistency loss.
Creates state-level supervision by matching T=2 intermediate states with T=1
final states when the T=2 output equals the T=1 input (string comparison only).
No arithmetic is computed - only evaluator-supplied values are compared.
"""
from __future__ import annotations
... | {
"id": "45c06e33-a5ec-4c2a-8271-c32bc6489ce8",
"created_at": "2026-08-14 15:17:57.895939+00:00",
"db_md5": "59ff1f01bde307f0fb0c532918a61cc3",
"submitter": "chad-atexpedient",
"github_login": "chad-atexpedient",
"run_id": "0dbfa6da-81f6-43fd-b4eb-68e34e343f4a",
"tier": "easy",
"dataset_id": "e5",
"st... | {
"score": {
"mean_loss": 2.295598248365886,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.009583333345750967
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.321414384072137,
"example_count": 600,... |
45c1cc68-bcbc-4145-9340-b55aafadcf80 | easy | Cree0618 | 2026-08-30 14:18:28.504648+00:00 | succeeded | 7b69b0dc9e34436cacb0b8dbfd76c5825c00b5ad6d166e467bfa9834deb5a6f1 | 8,815 | null | """E047: bit-serial NeuralHorner challenger for modular squaring."""
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
PAD, X, T, ANS, EOS, OFF = 0... | {
"id": "45c1cc68-bcbc-4145-9340-b55aafadcf80",
"created_at": "2026-08-30 14:18:28.504648+00:00",
"db_md5": "8be5f0c1dc7b23506427e46e09dcd33b",
"submitter": "Cree0618",
"github_login": "Cree0618",
"run_id": "07e44afb-ef21-4ede-bf23-20e19c8becd1",
"tier": "easy",
"dataset_id": "e5",
"status": "succeede... | {
"score": {
"mean_loss": 2.1875024568424215,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0045833333705862355
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.188093753139,
"example_count": 600,
... |
45c49b04-55f5-4a79-862e-c05d6f5dbf34 | easy | ArkinDharawat | 2026-08-13 17:24:29.173453+00:00 | succeeded | 71fb474a2ea2e25dc62b8c627199d72cc6c583a5ab4cfe954e579ec12c15b225 | 7,631 | 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": "45c49b04-55f5-4a79-862e-c05d6f5dbf34",
"created_at": "2026-08-13 17:24:29.173453+00:00",
"db_md5": "947b08db758f7de419fbe296519e29fb",
"submitter": "Arkin Dharawat",
"github_login": "ArkinDharawat",
"run_id": "e665be63-a7ef-490d-86f0-361ee96d45d4",
"tier": "easy",
"dataset_id": "e3",
"status"... | {
"score": {
"mean_loss": 2.2742468902681034,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.016250000121071934
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.3068413527813285,
"example_count": 80... |
45c707ed-c589-4b90-89b7-acde5a79873a | easy | 0Chris5R | 2026-08-02 16:25:48.851699+00:00 | succeeded | dc52d293c2f8167690b5b714d561a44eb641d4e444ad27af5b317235a74bad03 | 25,234 | 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": "45c707ed-c589-4b90-89b7-acde5a79873a",
"created_at": "2026-08-02 16:25:48.851699+00:00",
"db_md5": "89dcbede023eeb1de20690f87d97203e",
"submitter": "Chris ",
"github_login": "0Chris5R",
"run_id": "dbdfb9b1-a6e3-46e0-925f-1623d9dabd51",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.471585583058708,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.014166666666666668
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.299530320103393,
"example_count": 600,... |
45e59f70-d455-4e22-9d45-31276e17cc53 | easy | sapient-sapiens | 2026-08-13 19:36:28.736824+00:00 | succeeded | dcea86074792e314b7adf48ec0df7f89802d98843dedc7b4baffa0e6598fbf4d | 8,537 | null | """E3 fixed-T=2: one encoder plus a four-block tied state operator."""
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
W... | {
"id": "45e59f70-d455-4e22-9d45-31276e17cc53",
"created_at": "2026-08-13 19:36:28.736824+00:00",
"db_md5": "4b6c1af0d7ae59e75a3a9eea5ee7937a",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "a524f458-a11a-40df-8158-d9dbbcac736c",
"tier": "easy",
"dataset_id": "e3",
"status": "su... | {
"score": {
"mean_loss": 2.1259108832336295,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.00625
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.1234117115953164,
"example_count": 800,
... |
45e737c7-94d0-41ec-8747-64c3f42013ec | easy | DDanlov | 2026-08-13 22:42:32.136135+00:00 | failed | 5af9b439a1093cf32fb35da3d6dd93bde3bc6996adabf1ab2f6b96ef64bd71bf | 22,142 | null | import torch
import torch.nn as nn
import torch.nn.functional as F
import math
from dataclasses import dataclass
from typing import Optional
from torch.optim import Optimizer
try:
from client import Submission, assert_model_state, OptimizerBundle
except ImportError:
Submission = None
assert_model_state = N... | {
"id": "45e737c7-94d0-41ec-8747-64c3f42013ec",
"created_at": "2026-08-13 22:42:32.136135+00:00",
"db_md5": "5d4b2514861db2a9a5bbc7b69615ea8f",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "52f13984-2248-4747-be61-16feebdc3739",
"tier": "easy",
"dataset_id": "e3",
"status": "failed",
... | null
|
45ea7a40-bd7d-40db-a820-62399421d6ea | easy | DDanlov | 2026-08-09 00:22:55.380916+00:00 | succeeded | c04b6aba6b7323a48732d52f57fa1c83c421794f6ef9e165341f460b5b17e1bf | 15,978 | null | from __future__ import annotations
import sys
import os
import math
import time
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim.optimizer import Optimizer
VOCAB_SIZE = 17
DIGIT_OFFSET = 7
class RohanShampoo(Optimizer):
"""Self-contained RohanShampoo optimizer with eigenvalue ma... | {
"id": "45ea7a40-bd7d-40db-a820-62399421d6ea",
"created_at": "2026-08-09 00:22:55.380916+00:00",
"db_md5": "9faa334cf93ed0824d1fa303bbcf72d3",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "bbf6f702-25dd-401e-80bf-dd7e39664fd7",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 24.19670632096419,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0020833333333333333
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 27.811106365357812,
"example_count": 60... |
45ee068b-0780-45e7-8a3b-3cac0c67d44d | easy | erdavis0 | 2026-08-23 15:58:16.721462+00:00 | succeeded | d4e7b6f94c5a17f720147aa5b1ff3dedb9aca4341c22907ba39d97352562ee35 | 21,596 | null | """Learned confluent brickwork exchanges on field-aligned token tapes.
Opaque public token categories are routed into fixed right-aligned N, X, and T
tapes. Two learned-computation lanes then apply the same complete collection
of adjacent-place rewrites in opposite even/odd orders. One learned pair cell
is shared ov... | {
"id": "45ee068b-0780-45e7-8a3b-3cac0c67d44d",
"created_at": "2026-08-23 15:58:16.721462+00:00",
"db_md5": "e9eede9164a2380d5b1e85c703eae702",
"submitter": "Ethan",
"github_login": "erdavis0",
"run_id": "3c856979-a0d4-4ff7-81d2-316deba15f0d",
"tier": "easy",
"dataset_id": "e7",
"status": "succeeded",... | {
"score": {
"mean_loss": 9.49156141281128,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.01724598975852132
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 9.225957870483398,
"example_count": 85,
... |
45f4d4fe-8cca-4760-acf2-8cde01c51156 | easy | shirvani-jr | 2026-08-28 09:35:53.584813+00:00 | succeeded | 7098283b57b984c8ac1f008bc1b607e0cc041d272183ed624a6c39045e390914 | 48,997 | null | """Adaptive learned discrete recurrent fabric for One Layer Deeper.
A gradient-trained, operation-free recurrent computational machine. The
model learns, from the competition's endpoint supervision: how to parse the
prompt into a bank of categorical digit registers, a reusable transition
circuit applied serially, a l... | {
"id": "45f4d4fe-8cca-4760-acf2-8cde01c51156",
"created_at": "2026-08-28 09:35:53.584813+00:00",
"db_md5": "483a76b09649c8063e9493d02bdb3503",
"submitter": "Ali",
"github_login": "shirvani-jr",
"run_id": "8b7d175c-4d6d-4566-b786-9d5f31884622",
"tier": "easy",
"dataset_id": "e1",
"status": "succeeded"... | {
"score": {
"mean_loss": 1.8630589842796326,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.09333333000540733
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 1.818447232246399,
"example_count": 100,... |
45f61366-8b77-49b0-8382-ff3bffb7022d | easy | khushidahi | 2026-08-20 02:48:04.797534+00:00 | succeeded | ca43ace10976719cf903eb3a835d6c84323d6191bd6ce79df48b12e5fffccc53 | 26,878 | null | """R286: monic-residual parallel quadratic functional root.
This candidate is deliberately diagnostic. It trains on evaluator-owned final
labels at several depths so that one shared transition must compose. The model
never converts N or X to native numbers and never implements multiplication or
modular reduction.
T... | {
"id": "45f61366-8b77-49b0-8382-ff3bffb7022d",
"created_at": "2026-08-20 02:48:04.797534+00:00",
"db_md5": "33c7617319925b4f7d09425300d0ca05",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "bbbff087-f5df-430c-ba6c-fb71f4779e6d",
"tier": "easy",
"dataset_id": "e3",
"status": "succ... | {
"score": {
"mean_loss": 3.0176320799526133,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.008750000018626452
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.9571668088661807,
"example_count": 80... |
46022ec5-2730-40d6-8318-d0ca28659d37 | easy | khushidahi | 2026-08-21 22:18:50.414538+00:00 | succeeded | 4908ba1d471e6df03fa2fc07325f8a5bb4bcba2aacf42bda0152acac442aca29 | 36,481 | null | """R380: late short-operand polish for the rank-eight polynomial root.
This reverses R318's optimization order. The first 75% fits the supplied
endpoint distribution; Adam state is then reset and the final 25% emphasizes
short operands immediately before evaluation. No target or forward mechanism
changes. This tests w... | {
"id": "46022ec5-2730-40d6-8318-d0ca28659d37",
"created_at": "2026-08-21 22:18:50.414538+00:00",
"db_md5": "a78979160ef6404362c355bb2ef97c00",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "3175fe12-3c4c-4157-a5b9-5ad6f197250a",
"tier": "easy",
"dataset_id": "e3",
"status": "succ... | {
"score": {
"mean_loss": 2.9111490805443925,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004375
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.9029080726341085,
"example_count": 800,
... |
460a3c5c-2167-45c8-ae35-c7e64adb2a12 | easy | sirish-gambhira | 2026-08-09 19:45:03.590374+00:00 | succeeded | 35cd38665998e6ff715d4dc22a08a946d7517294a9319fd1fc64df9a0eaecfe9 | 7,812 | null | """Slim decimal-structure-aware Transformer for token arithmetic."""
from __future__ import annotations
import math
import torch
import torch.nn.functional as F
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
assert_... | {
"id": "460a3c5c-2167-45c8-ae35-c7e64adb2a12",
"created_at": "2026-08-09 19:45:03.590374+00:00",
"db_md5": "bd725b2d3b9272053cd8d5a18a28b022",
"submitter": "Sirish Gambhira",
"github_login": "sirish-gambhira",
"run_id": "e8747525-27bd-4c24-b647-4377cc9ad4f6",
"tier": "easy",
"dataset_id": "e5",
"stat... | {
"score": {
"mean_loss": 4.660617828369141,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.005000000121071935
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.563220977783203,
"example_count": 600,... |
461ab882-fdc7-4787-813a-623404fa35f7 | easy | khushidahi | 2026-08-09 02:20:27.786668+00:00 | succeeded | b6825ce7526749bd5b8238b2ad08ac3495cebf89dc61a573a318f7a7ebd05bd8 | 12,576 | null | """R45 full-budget, learned-symbol baseline.
This model intentionally avoids task-specific numerical decoding:
- no fixed digit values;
- no decimal-to-integer conversion;
- no FoNE parser;
- no solver-derived supervision.
The evaluator time allowance determines the LR schedule horizon. `max_steps`
is deliberately hi... | {
"id": "461ab882-fdc7-4787-813a-623404fa35f7",
"created_at": "2026-08-09 02:20:27.786668+00:00",
"db_md5": "030ea406eca7d9b2e115fe13514c751a",
"submitter": "khushidahi",
"github_login": "khushidahi",
"run_id": "4510c345-ce50-4834-bb25-d522069c2405",
"tier": "easy",
"dataset_id": "e5",
"status": "succ... | {
"score": {
"mean_loss": 2.1771983661361642,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.0037500000403573115
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.182116746902466,
"example_count": 60... |
46239027-964b-4438-8b5b-e4697c406f29 | easy | DDanlov | 2026-08-15 14:11:53.437355+00:00 | succeeded | 2be29914cf149f7332bde5d60956a02f6bdac791177652639cbe669f8c4562ca | 18,985 | null | """
Official Submission for One Layer Deeper Challenge
Architecture: Pure No-Norm, No-Skip DKS Recurrent Core with Double Embedding (dim=512),
Input RMS Normalization only, 64-Channel Harmonic+Dyadic Torus, Power-of-10 Alignment, and RohanShampoo.
"""
from __future__ import annotations
import math
from typing import O... | {
"id": "46239027-964b-4438-8b5b-e4697c406f29",
"created_at": "2026-08-15 14:11:53.437355+00:00",
"db_md5": "523064fce00745db1dbae9272bf92154",
"submitter": "DDanlov",
"github_login": "DDanlov",
"run_id": "3b034c7e-80d5-4e90-8761-31f3c8fd8fa0",
"tier": "easy",
"dataset_id": "e5",
"status": "succeeded"... | {
"score": {
"mean_loss": 2.334067072792929,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.004583333333333333
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.287222665655239,
"example_count": 600,... |
46289a22-fced-4f1f-8421-b26f04359668 | easy | sapient-sapiens | 2026-08-18 13:36:48.703914+00:00 | succeeded | 3ff9fc5a0835b3363ca4ae57f78751bc4bdf6a86e39ef66bf5ab39ca3bbe5efe | 29,237 | 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": "46289a22-fced-4f1f-8421-b26f04359668",
"created_at": "2026-08-18 13:36:48.703914+00:00",
"db_md5": "50638b218fe7656fd6b8e87e4cee6538",
"submitter": "Ertondy",
"github_login": "sapient-sapiens",
"run_id": "732ff7d2-46fa-4d59-b467-d54d7de5d1ee",
"tier": "easy",
"dataset_id": "e8",
"status": "su... | {
"score": {
"mean_loss": 3.541436791419983,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.05254010856151581
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 3.551825761795044,
"example_count": 85,
... |
4629a9b5-d0f9-4e06-b83f-d1a732c1b565 | easy | shreyash-chonkie | 2026-09-01 02:02:21.217241+00:00 | succeeded | 3b5ef230531ee5171e25b472ba0a302cc3e9417afb831668e4c996624534252a | 16,171 | null | """Learned token scanner feeding a random recurrent path reservoir."""
from __future__ import annotations
import torch
from torch import Tensor, nn
from benchmark import (
ModelSpec,
OptimizerBundle,
OptimizerSpec,
Submission,
TokenLossBatch,
assert_model_state,
)
NUM_OPERANDS = 4
NUM_OPERA... | {
"id": "4629a9b5-d0f9-4e06-b83f-d1a732c1b565",
"created_at": "2026-09-01 02:02:21.217241+00:00",
"db_md5": "3cd45c0013aee428ae24305092d98814",
"submitter": "Shreyash",
"github_login": "shreyash-chonkie",
"run_id": "d2845a8d-91fb-47dd-b2d2-a950a3a21df7",
"tier": "easy",
"dataset_id": "e1",
"status": "... | {
"score": {
"mean_loss": 6.346650494037931e-08,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 1.0
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 6.182869327631124e-08,
"example_count": 100,
... |
462ccd56-a552-43b5-bb54-08628f0ceadd | easy | KaustubhKumar05 | 2026-08-27 11:52:04.905375+00:00 | succeeded | dbcc71c4c0a38e23cc4e64c5be18cc668bb50ce47a094b8b5436f5fa84606a14 | 14,044 | null | """abacus_gromov — Gromov-style: NO attention, plain MLP, MSE loss, no regularisation.
§62: Gromov 2301.02679 groks modular arithmetic with a TWO-LAYER FULLY-CONNECTED
net, vanilla GD, MSE loss and NO regularisation, and derives closed-form weights.
His actual architecture does NOT port: it one-hot encodes operands o... | {
"id": "462ccd56-a552-43b5-bb54-08628f0ceadd",
"created_at": "2026-08-27 11:52:04.905375+00:00",
"db_md5": "cfe87322d620a601e25a918010f5954e",
"submitter": "koz",
"github_login": "KaustubhKumar05",
"run_id": "3e0808bd-21e4-43d9-8216-951c7676e24f",
"tier": "easy",
"dataset_id": "e3",
"status": "succee... | {
"score": {
"mean_loss": 2.749796565858408,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.003125
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 2.7500404826121314,
"example_count": 800,
... |
4634aeb1-b75c-4a6d-8067-e4da4d791e80 | easy | chad-atexpedient | 2026-08-05 02:22:40.230891+00:00 | succeeded | c5955d5e44096b71de0ac143bce2ebf3700a6798069b7979d5abdd4807f96962 | 4,481 | null | """R023: batch-2048 one-pass RoPE model with learning rate 1.2e-3."""
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 = 128
HEA... | {
"id": "4634aeb1-b75c-4a6d-8067-e4da4d791e80",
"created_at": "2026-08-05 02:22:40.230891+00:00",
"db_md5": "79efe62206a4f9211e174c1fbd69287b",
"submitter": "chad-atexpedient",
"github_login": "chad-atexpedient",
"run_id": "7575f563-d6e4-47e9-939c-71c466bd243b",
"tier": "easy",
"dataset_id": "e5",
"st... | {
"score": {
"mean_loss": 3.330697536468506,
"primary_metric": "mean_exact_accuracy",
"num_measurements": 2,
"mean_exact_accuracy": 0.006666666828095913
},
"seeds": [
{
"seed": 74,
"evaluation": {
"ood": {
"loss": 4.073378562927246,
"example_count": 600,... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.