SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2106.06770 · 2021

What can linearized neural networks actually say about generalization?

arXiv · PDF · Open in the Atlas

Code that ran

3 of 4 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, 0 violated it and 3 have no contract verdict; 0 failures of our instrument; 1 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; 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/2106.06770.json · badge: /api/ran/2106.06770.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
gortizji/linearized-networks canonical (classifier label)
classifier basis: census
3 ran · 1 no recorded run
FunctionStatusWhere it lives
binary_cross_entropy_loss_with_logits Ran
ran on a synthesized input
gortizji/linearized-networks/utils/loss.py
code served (permissive licence) · get_code("a97ccf4556b4a717")
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 ArrayImpl, 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 jax.numpy as jnp
import optax

# Create small test tensors
batch_size = 2
num_features = 4

# logits: shape (batch_size, num_features) - raw model outputs
logits = jnp.array([
    [0.1, -0.2, 0.3, -0.1],
    [-0.5, 0.2, -0.3, 0.4]
], dtype=jnp.float32)

# labels: shape (batch_size,) - binary labels (0 or 1)
labels = jnp.array([1, 0], dtype=jnp.float32)

ARGS = {
    "logits": logits,
    "labels": labels,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, optax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, 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
mse_loss Ran
ran on a synthesized input
gortizji/linearized-networks/utils/loss.py
code served (permissive licence) · get_code("82c402b4b47c433b")
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 ArrayImpl, 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 jax.numpy as jnp
import optax

def mse_loss(preds, labels):
    return jnp.mean(optax.l2_loss(preds, labels))

# Create small test tensors with distinct dimensions for clarity
batch_size = 2
output_dim = 4

# preds: model predictions, shape (batch_size, output_dim)
preds = jnp.array([[1.0, 2.0, 3.0, 4.0],
                    [5.0, 6.0, 7.0, 8.0]])

# labels: ground truth labels, shape (batch_size, output_dim)
labels = jnp.array([[1.1, 2.1, 3.1, 4.1],
                     [5.1, 6.1, 7.1, 8.1]])

ARGS = {
    "preds": preds,
    "labels": labels
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, optax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, 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
softmax_cross_entropy_loss Ran
ran on a synthesized input
gortizji/linearized-networks/utils/loss.py
code served (permissive licence) · get_code("581398948f76cfae")
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 ArrayImpl, 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 jax
import jax.numpy as jnp
import optax

# Set seed for reproducibility
import numpy as np
np.random.seed(0)

# Create small test data
batch_size = 2
num_classes = 4

# Create logits: shape (batch_size, num_classes)
logits = jnp.array([
    [1.0, 2.0, 0.5, 1.5],
    [0.2, 1.5, 2.0, 0.8]
], dtype=jnp.float32)

# Create labels as integer class indices: shape (batch_size,)
labels = jnp.array([1, 2], dtype=jnp.int32)

ARGS = {
    "logits": logits,
    "labels": labels,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, numpy, optax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, 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
mixed_derivative_nad_decomposition Not yet run
unverified (harvested, no recorded run)
gortizji/linearized-networks/neural_kernels/nads.py
code served (permissive licence) · get_code("58979551de01a0e4")
How it ran

no invocation recorded.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 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.

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

For certain infinitely-wide neural networks, the neural tangent kernel (NTK) theory fully characterizes generalization, but for the networks used in practice, the empirical NTK only provides a rough first-order approximation. Still, a growing body of work keeps leveraging this approximation to successfully analyze important deep learning phenomena and design algorithms for new applications. In our work, we provide strong empirical evidence to determine the practical validity of such approximation by conducting a systematic comparison of the behavior of different neural networks and their linear approximations on different tasks. We show that the linear approximations can indeed rank the learning complexity of certain tasks for neural networks, even when they achieve very different performances. However, in contrast to what was previously reported, we discover that neural networks do not always perform better than their kernel approximations, and reveal that the performance gap heavily depends on architecture, dataset size and training task. We discover that networks overfit to these tasks mostly due to the evolution of their kernel during training, thus, revealing a new type of implicit bias.

For agents

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

get_harvested_code_for_paper("2106.06770")
get_code_for_paper("2106.06770")
have("2106.06770")

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

curl https://syntology.ai/api/ran/2106.06770.json

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

[![Syntology run record](https://syntology.ai/api/ran/2106.06770.svg)](https://syntology.ai/paper/2106.06770)

Connect an agent — have() is free.