SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2209.03695 · NeurIPS · 2022

Training Scale-Invariant Neural Networks on the Sphere Can Happen in Three Regimes

Ekaterina Lobacheva, Dmitry Vetrov, Maxim Kodryan, Maksim Nakhodnov

arXiv · PDF · Open in the Atlas

Code that ran

11 of 15 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 1 violated it and 8 have no contract verdict; 2 failures of our instrument; 4 no recorded run), measured 2026-11-22; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2209.03695.json · badge: /api/ran/2209.03695.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
tipt0p/three_regimes_on_the_sphere canonical (classifier label)
classifier basis not recorded on the link
11 ran · 4 no recorded run
FunctionStatusWhere it lives
ConvBNrelu Ran
ran on a synthesized input
tipt0p/three_regimes_on_the_sphere/nets/cifar_net.py
code served (permissive licence) · get_code("a164b6cd82f40a9f")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Sequential. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    "in_channels": 3,
    "out_channels": 16,
    "stride": 1
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Sequential — a Sequential output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Sequential output the recorded ndim 1 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
Lop Ran
ran; our contract draft was wrong, not the code
tipt0p/three_regimes_on_the_sphere/get_info_funcs.py
code served (permissive licence) · get_code("19b54a137a0bbdb7")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 3, dtype family float; no output shape is stated. It ran in syntology-harvest-sandbox:v5-deps (sha256:6a292aa1aed1) at 2026-09-27T00:58:14Z. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create input tensors
# ys: output of a computation that requires grad
xs = torch.tensor([[1.0, 2.0, 3.0, 4.0],
                    [5.0, 6.0, 7.0, 8.0]], requires_grad=True)

# Create a simple computation graph
ys = (xs ** 2).sum()

# ws: gradient outputs weights, should match shape of ys
# Since ys is a scalar, ws should be a scalar
ws = torch.tensor(1.0)

ARGS = {
    "ys": ys,
    "xs": xs,
    "ws": ws
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 3, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image syntology-harvest-sandbox:v5-deps (sha256:6a292aa1aed1); not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
block Ran
ran on a synthesized input
tipt0p/three_regimes_on_the_sphere/nets/convnet.py
code served (permissive licence) · get_code("e9d3b739701f64b2")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    "input": 3,
    "output": 16
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
block Ran
ran on a synthesized input
tipt0p/three_regimes_on_the_sphere/nets/convnet_si.py
code served (permissive licence) · get_code("f9e345905349850a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# The function takes two parameters: input and output
# These are integers representing the number of input and output channels
# for the Conv2d layer that's created inside the function

input_channels = 3
output_channels = 16

ARGS = {
    "input": input_channels,
    "output": output_channels,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
block Ran
ran on a synthesized input
tipt0p/three_regimes_on_the_sphere/nets/convnet_si_af.py
code served (permissive licence) · get_code("bf9fa9508b358457")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# The function takes two parameters: input and output
# These are integers representing the number of input and output channels
# for the Conv2d layer at the start of the block

input_channels = 3
output_channels = 16

ARGS = {
    "input": input_channels,
    "output": output_channels,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
block Ran
ran on a synthesized input
tipt0p/three_regimes_on_the_sphere/nets/convnet_nobn.py
code served (permissive licence) · get_code("a37fca5fb609181f")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    "input": 3,
    "output": 16
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
calc_cos Ran
ran on a synthesized input
tipt0p/three_regimes_on_the_sphere/get_info.py
code served (permissive licence) · get_code("257b3850722caed2")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned float, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# The function iterates over params_1 and params_2 with zip
# It calls .sum().item() on tensors, so we need iterables of tensors
# We'll create two lists of small tensors

params_1 = [
    torch.randn(2, 4),
    torch.randn(3, 5),
]

params_2 = [
    torch.randn(2, 4),
    torch.randn(3, 5),
]

ARGS = {
    "params_1": params_1,
    "params_2": params_2,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
calc_l2 Ran
ran on a synthesized input
tipt0p/three_regimes_on_the_sphere/get_info.py
code served (permissive licence) · get_code("c67965018c758260")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned float, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# Create two small parameter lists
# Each parameter should be a tensor that supports subtraction, squaring, sum, and item()
params_1 = [
    torch.randn(4, 3),
    torch.randn(2, 5),
    torch.randn(3)
]

params_2 = [
    torch.randn(4, 3),
    torch.randn(2, 5),
    torch.randn(3)
]

ARGS = {
    "params_1": params_1,
    "params_2": params_2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
check_si_name Ran
ran, violated the contract we drafted
tipt0p/three_regimes_on_the_sphere/nets/convnet_si.py
code served (permissive licence) · get_code("d2c03e2aaa143601")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned bool, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

ARGS = {
    'n': 'conv1',
    'model_name': 'ResNet18'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type bool, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
check_si_name Ran
ran on a synthesized input
tipt0p/three_regimes_on_the_sphere/get_info.py
code served (permissive licence) · get_code("4a4f943527362329")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned bool, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# The function check_si_name takes:
# - n: a string (parameter name to check)
# - model_name: a string specifying the model type (default 'ResNet18')
#
# The function only uses string operations (checking if substrings are in n)
# and conditional logic based on model_name. It doesn't actually use a model object,
# just the model_name string for routing logic.

ARGS = {
    'n': 'conv1.weight',
    'model_name': 'ResNet18'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type bool, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
unflatten_like Ran
ran; our contract draft was wrong, not the code
tipt0p/three_regimes_on_the_sphere/get_info_funcs.py
code served (permissive licence) · get_code("6d8daee3396e21d5")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# Create a flattened vector
vector = torch.randn(15)

# Create a list of tensors with different shapes to unflatten like
likeTensorList = [
    torch.randn(2, 3),      # 6 elements
    torch.randn(4, 2),      # 8 elements
    torch.randn(1),         # 1 element
]

ARGS = {
    "vector": vector,
    "likeTensorList": likeTensorList,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
Rop Not yet run
unverified (harvested, no recorded run)
tipt0p/three_regimes_on_the_sphere/get_info_funcs.py
code served (permissive licence) · get_code("8bc49ef2a0810046")
How it ran

no invocation recorded.

get_model_extra Not yet run
unverified (harvested, no recorded run)
tipt0p/three_regimes_on_the_sphere/nb_utils.py
code served (permissive licence) · get_code("157cf0b88320c24e")
How it ran

no invocation recorded.

get_model_str Not yet run
unverified (harvested, no recorded run)
tipt0p/three_regimes_on_the_sphere/nb_utils.py
code served (permissive licence) · get_code("c415f0e6392ad2af")
How it ran

no invocation recorded.

svhn_loaders Not yet run
unverified (harvested, no recorded run)
tipt0p/three_regimes_on_the_sphere/data.py
code served (permissive licence) · get_code("182de3ac9284f86e")
How it ran

no invocation recorded.

Also present in other papers' code

Of 15 samples lifted from this paper's linked repositories, 15 carry a function identity and 0 do not. Of the 15 distinct functions with an identity, 11 are also present in other papers whose code we hold (11 with at least one paper dated earlier, 0 the same month, 7 later) and 4 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
unflatten_like 3 other papers
1 dated earlier, 2 later
2106.15739 (2021-06)
On the Periodic Behavior of Neural Network Training with Batch Normalization and Weight Decay
2312.17173 (2023-12)
Non-Vacuous Generalization Bounds for Large Language Models
2410.22113 (2024-10)
Where Do Large Learning Rates Lead Us?
ConvBNrelu 2 other papers
2 dated earlier
2003.02139 (2020-03)
Rethinking Parameter Counting in Deep Models: Effective Dimensionality Revisited
2106.15739 (2021-06)
On the Periodic Behavior of Neural Network Training with Batch Normalization and Weight Decay
Lop 2 other papers
1 dated earlier, 1 later
2106.15739 (2021-06)
On the Periodic Behavior of Neural Network Training with Batch Normalization and Weight Decay
2410.22113 (2024-10)
Where Do Large Learning Rates Lead Us?
Rop 2 other papers
1 dated earlier, 1 later
2106.15739 (2021-06)
On the Periodic Behavior of Neural Network Training with Batch Normalization and Weight Decay
2410.22113 (2024-10)
Where Do Large Learning Rates Lead Us?
block 2 other papers
2 dated earlier
2003.02139 (2020-03)
Rethinking Parameter Counting in Deep Models: Effective Dimensionality Revisited
2106.15739 (2021-06)
On the Periodic Behavior of Neural Network Training with Batch Normalization and Weight Decay
check_si_name 2 other papers
1 dated earlier, 1 later
2106.15739 (2021-06)
On the Periodic Behavior of Neural Network Training with Batch Normalization and Weight Decay
2410.22113 (2024-10)
Where Do Large Learning Rates Lead Us?
get_model_extra 2 other papers
1 dated earlier, 1 later
2106.15739 (2021-06)
On the Periodic Behavior of Neural Network Training with Batch Normalization and Weight Decay
2410.22113 (2024-10)
Where Do Large Learning Rates Lead Us?
get_model_str 2 other papers
1 dated earlier, 1 later
2106.15739 (2021-06)
On the Periodic Behavior of Neural Network Training with Batch Normalization and Weight Decay
2410.22113 (2024-10)
Where Do Large Learning Rates Lead Us?
svhn_loaders 2 other papers
1 dated earlier, 1 later
2106.15739 (2021-06)
On the Periodic Behavior of Neural Network Training with Batch Normalization and Weight Decay
2410.22113 (2024-10)
Where Do Large Learning Rates Lead Us?
block 1 other paper
1 dated earlier
2106.15739 (2021-06)
On the Periodic Behavior of Neural Network Training with Batch Normalization and Weight Decay
block 1 other paper
1 dated earlier
2106.15739 (2021-06)
On the Periodic Behavior of Neural Network Training with Batch Normalization and Weight Decay

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

A fundamental property of deep learning normalization techniques, such as batch normalization, is making the pre-normalization parameters scale invariant. The intrinsic domain of such parameters is the unit sphere, and therefore their gradient optimization dynamics can be represented via spherical optimization with varying effective learning rate (ELR), which was studied previously. However, the varying ELR may obscure certain characteristics of the intrinsic loss landscape structure. In this work, we investigate the properties of training scale-invariant neural networks directly on the sphere using a fixed ELR. We discover three regimes of such training depending on the ELR value: convergence, chaotic equilibrium, and divergence. We study these regimes in detail both on a theoretical examination of a toy example and on a thorough empirical analysis of real scale-invariant deep learning models. Each regime has unique features and reflects specific properties of the intrinsic loss landscape, some of which have strong parallels with previous research on both regular and scale-invariant neural networks training. Finally, we demonstrate how the discovered regimes are reflected in conventional training of normalized networks and how they can be leveraged to achieve better optima.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2209.03695")
get_code_for_paper("2209.03695")
have("2209.03695")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2209.03695.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2209.03695.svg)](https://syntology.ai/paper/2209.03695)

Connect an agent — have() is free.