SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2412.09115 · ICLR · 2025

Vision CNNs trained to estimate spatial latents learned similar ventral-stream-aligned representations

Joshua Tenenbaum, Jeremy Schwartz, James Dicarlo, Yudi Xie, Weichen Huang, Esther Alter

arXiv · PDF · Open in the Atlas

Code that ran

4 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 0 have no contract verdict; 4 failures of our instrument; 0 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/2412.09115.json · badge: /api/ran/2412.09115.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
YudiXie/multitask-vision canonical (classifier label)
classifier basis: census
4 ran · 0 no recorded run
FunctionStatusWhere it lives
annotate_heatmap Ran
ran; our contract draft was wrong, not the code
YudiXie/multitask-vision/analysis.py
code served (permissive licence) · get_code("c59639a89e782b9f")
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 numpy as np
import matplotlib.pyplot as plt
import matplotlib.ticker

# Create a minimal figure and axes
fig, ax = plt.subplots()

# Create sample data for the heatmap
data = np.array([[1.0, 2.0, 3.0],
                 [4.0, 5.0, 6.0],
                 [7.0, 8.0, 9.0]])

# Create an AxesImage object using imshow
im = ax.imshow(data, cmap='viridis')

# Define the arguments for annotate_heatmap
ARGS = {
    'im': im,
    'data': data,
    'valfmt': '{x:.2f}',
    'textcolors': ('black', 'white'),
    'threshold': None,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
matplotlib, numpy — 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
append_activations Ran
ran; our contract draft was wrong, not the code
YudiXie/multitask-vision/activity.py
code served (permissive licence) · get_code("e9d77a401b4f3051")
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 function. 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
from collections import defaultdict

torch.manual_seed(0)

# Create a minimal activity_dict that is a defaultdict with list factory
activity_dict = defaultdict(list)

# Create a name string for the module
name = "conv_layer"

# Define ARGS mapping all parameters of append_activations
ARGS = {
    "name": name,
    "activity_dict": activity_dict,
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
get_stat_str Ran
ran; our contract draft was wrong, not the code
YudiXie/multitask-vision/analysis.py
code served (permissive licence) · get_code("e68de799c4342ea9")
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 str. 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 numpy as np

torch.manual_seed(0)

# Analyze the function:
# - p_value: a numeric value (float) that will be compared with p (0.05, 0.005, 0.0005, etc.)
# - maxasterix: optional int parameter that limits the number of asterisks

# Test case 1: p_value that produces asterisks
ARGS = {
    'p_value': 0.001,
    'maxasterix': 5
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
heatmap Ran
ran; our contract draft was wrong, not the code
YudiXie/multitask-vision/analysis.py
code served (permissive licence) · get_code("8007d64c247ab34d")
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 numpy as np
from matplotlib import pyplot as plt

# Create a small 2D numpy array for the heatmap data
data = np.random.rand(4, 3)

# Create row and column labels
row_labels = ['Row_A', 'Row_B', 'Row_C', 'Row_D']
col_labels = ['Col_1', 'Col_2', 'Col_3']

# Create a matplotlib figure and axes
fig, ax = plt.subplots()

# Create colorbar keyword arguments dictionary
cbar_kw = {'shrink': 0.8}

# Define the cbarlabel
cbarlabel = 'Intensity'

# Additional kwargs for imshow (e.g., cmap)
imshow_kwargs = {'cmap': 'viridis'}

ARGS = {
    'data': data,
    'row_labels': row_labels,
    'col_labels': col_labels,
    'ax': ax,
    'cbar_kw': cbar_kw,
    'cbarlabel': cbarlabel,
}

# Merge imshow_kwargs into ARGS
ARGS.update(imshow_kwargs)
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
matplotlib, numpy — 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

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, 1 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 0 later) and 3 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
annotate_heatmap 1 other paper
1 dated earlier
2207.00735 (2022-07)
Can Language Models Make Fun? A Case Study in Chinese Comical Crosstalk

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

Studies of the functional role of the primate ventral visual stream have traditionally focused on object categorization, often ignoring -despite much prior evidence -its role in estimating "spatial" latents such as object position and pose. Most leading ventral stream models are derived by optimizing networks for object categorization, which seems to imply that the ventral stream is also derived under such an objective. Here, we explore an alternative hypothesis: Might the ventral stream be optimized for estimating spatial latents? And a closely related question: How different -if at all -are representations learned from spatial latent estimation compared to categorization? To ask these questions, we leveraged synthetic image datasets generated by a 3D graphic engine and trained convolutional neural networks (CNNs) to estimate different combinations of spatial and category latents. We found that models trained to estimate just a few spatial latents achieve neural alignment scores comparable to those trained on hundreds of categories, and the spatial latent performance of models strongly correlates with their neural alignment. Spatial latent and category-trained models have very similar -but not identical -internal representations, especially in their early and middle layers. We provide evidence that this convergence is partly driven by non-target latent variability in the training data, which facilitates the implicit learning of representations of those non-target latents. Taken together, these results suggest that many training objectives, such as spatial latents, can lead to similar models aligned neurally with the ventral stream. Thus, one should not assume that the ventral stream is optimized for object categorization only. As a field, we need to continue to sharpen our measures of comparing models to brains to better understand the functional roles of the ventral stream.

For agents

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

get_harvested_code_for_paper("2412.09115")
get_code_for_paper("2412.09115")
have("2412.09115")

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

curl https://syntology.ai/api/ran/2412.09115.json

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

[![Syntology run record](https://syntology.ai/api/ran/2412.09115.svg)](https://syntology.ai/paper/2412.09115)

Connect an agent — have() is free.