SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2006.14331 · 2020

A Theoretical Framework for Target Propagation

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 2026-11-21; 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/2006.14331.json · badge: /api/ran/2006.14331.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
AlexanderMeulemans/TheoreticalFrameworkForTargetPropagation canonical (classifier label)
classifier basis: census
3 ran · 1 no recorded run
FunctionStatusWhere it lives
accuracy Ran
ran on a synthesized input
AlexanderMeulemans/TheoreticalFrameworkForTargetPropagation/lib/utils.py
code served (permissive licence) · get_code("d38d8a0b7b053904")
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

torch.manual_seed(0)

# Create predictions: batch_size=2, num_classes=4
# torch.max will select the argmax along dimension 1
predictions = torch.randn(2, 4)

# Create labels: batch_size=2, with class indices in range [0, 3]
labels = torch.tensor([1, 3])

ARGS = {
    "predictions": predictions,
    "labels": labels
}
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
choose_forward_optimizer Ran
ran on a synthesized input
AlexanderMeulemans/TheoreticalFrameworkForTargetPropagation/lib/utils.py
code served (permissive licence) · get_code("b256dc876fd34c06")
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 Adam. 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 minimal neural network that implements the methods called
class MinimalNet(nn.Module):
    def __init__(self):
        super().__init__()
        self.layers = nn.ModuleList([
            nn.Linear(4, 8),
            nn.Linear(8, 6),
            nn.Linear(6, 3)
        ])
    
    def forward(self, x):
        x = self.layers[0](x)
        x = self.layers[1](x)
        x = self.layers[2](x)
        return x
    
    def get_reduced_forward_parameter_list(self):
        # Return parameters excluding output layer
        params = list(self.layers[0].parameters())
        for layer in self.layers[1:-1]:
            params.extend(layer.parameters())
        return params
    
    def get_forward_parameter_list(self):
        # Return all parameters
        return self.parameters()
    
    def get_forward_parameter_list_first_layer(self):
        # Return only first layer parameters
        return self.layers[0].parameters()
    
    def get_forward_parameters_last_two_layers(self):
        # Return last two layers parameters
        params = list(self.layers[-2].parameters())
        params.extend(self.layers[-1].parameters())
        return params
    
    def get_forward_parameters_last_three_layers(self):
        # Return last three layers parameters
        params = list(self.layers[-3].parameters())
        params.extend(self.layers[-2].parameters())
        params.extend(self.layers[-1].parameters())
        return params
    
    def get_forward_parameters_last_four_layers(self):
        # Return all parameters (only have 3 layers)
        return self.parameters()

# Create args object with all required attributes
class Args:
    def __init__(self):
        self.freeze_BPlayers = False
        self.network_type = 'other'  # Not 'BP' or 'BPConv'
        self.shallow_training = False
        self.only_train_first_layer = False
        self.only_train_last_two_layers = False
        self.only_train_last_three_layers = False
        self.only_train_last_four_layers = False
        self.freeze_output_layer = False
        self.optimizer = 'Adam'
        self.lr = 0.001
        self.momentum = 0.9
        self.forward_wd = 0.0001
        self.beta1 = 0.9
        self.beta2 = 0.999
        self.epsilon = 1e-8

ARGS = {
    'args': Args(),
    'net': MinimalNet()
}
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 Adam — a Adam output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Adam output the recorded ndim 0 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
choose_optimizer Ran
ran on a synthesized input
AlexanderMeulemans/TheoreticalFrameworkForTargetPropagation/lib/utils.py
code served (permissive licence) · get_code("da6c0d07f0a5cfc2")
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 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
import numpy as np

torch.manual_seed(0)

class SimpleNet(nn.Module):
    def __init__(self):
        super().__init__()
        self.layer1 = nn.Linear(4, 3)
        self.layer2 = nn.Linear(3, 2)
        self.layers = [self.layer1, self.layer2]
    
    def get_forward_parameter_list(self):
        return list(self.parameters())
    
    def get_reduced_forward_parameter_list(self):
        return list(self.layer1.parameters())
    
    def get_forward_parameter_list_first_layer(self):
        return list(self.layer1.parameters())
    
    def get_forward_parameters_last_two_layers(self):
        return list(self.layer1.parameters()) + list(self.layer2.parameters())
    
    def get_forward_parameters_last_three_layers(self):
        return list(self.parameters())
    
    def get_forward_parameters_last_four_layers(self):
        return list(self.parameters())
    
    def get_feedback_parameter_list(self):
        return list(self.parameters())

class Args:
    def __init__(self):
        self.freeze_BPlayers = False
        self.network_type = 'MLP'
        self.shallow_training = False
        self.only_train_first_layer = False
        self.only_train_last_two_layers = False
        self.only_train_last_three_layers = False
        self.only_train_last_four_layers = False
        self.freeze_output_layer = False
        self.optimizer = 'Adam'
        self.optimizer_fb = 'Adam'
        self.lr = 0.001
        self.lr_fb = 0.001
        self.momentum = 0.9
        self.forward_wd = 0.0
        self.feedback_wd = 0.0
        self.beta1 = 0.9
        self.beta2 = 0.999
        self.beta1_fb = 0.9
        self.beta2_fb = 0.999
        self.epsilon = 1e-8
        self.epsilon_fb = 1e-8

args = Args()
net = SimpleNet()

ARGS = {
    'args': args,
    'net': net
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, 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 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
generate_data_from_teacher_network Not yet run
unverified (harvested, no recorded run)
AlexanderMeulemans/TheoreticalFrameworkForTargetPropagation/lib/builders.py
code served (permissive licence) · get_code("c3f6ba0f418c116c")
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 2026-11-21.

Repositories linked to this paper

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

Abstract

The success of deep learning, a brain-inspired form of AI, has sparked interest in understanding how the brain could similarly learn across multiple layers of neurons. However, the majority of biologically-plausible learning algorithms have not yet reached the performance of backpropagation (BP), nor are they built on strong theoretical foundations. Here, we analyze target propagation (TP), a popular but not yet fully understood alternative to BP, from the standpoint of mathematical optimization. Our theory shows that TP is closely related to Gauss-Newton optimization and thus substantially differs from BP. Furthermore, our analysis reveals a fundamental limitation of difference target propagation (DTP), a well-known variant of TP, in the realistic scenario of non-invertible neural networks. We provide a first solution to this problem through a novel reconstruction loss that improves feedback weight training, while simultaneously introducing architectural flexibility by allowing for direct feedback connections from the output to each hidden layer. Our theory is corroborated by experimental results that show significant improvements in performance and in the alignment of forward weight updates with loss gradients, compared to DTP.

For agents

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

get_harvested_code_for_paper("2006.14331")
get_code_for_paper("2006.14331")
have("2006.14331")

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

curl https://syntology.ai/api/ran/2006.14331.json

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

[![Syntology run record](https://syntology.ai/api/ran/2006.14331.svg)](https://syntology.ai/paper/2006.14331)

Connect an agent — have() is free.