SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2501.03992 · ICCV · 2025

NeuralSVG: An Implicit Representation for Text-to-Vector Generation

Daniel Cohen-Or, Yael Vinker, Elad Richardson, Sagi Polaczek

arXiv · PDF · Open in the Atlas

Code that ran

3 of 5 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; 2 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/2501.03992.json · badge: /api/ran/2501.03992.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
SagiPolaczek/NeuralSVG community (classifier label)
classifier basis not recorded on the link
3 ran · 2 no recorded run
FunctionStatusWhere it lives
calculate_ratio Ran
ran on a synthesized input
SagiPolaczek/NeuralSVG/src/models/nerf_mlp_multi.py
code served (permissive licence) · get_code("4a1b70be64371746")
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 numpy as np

# The function takes a ratio_string parameter and parses it
# It expects a string in the format "x:y" where x and y are numeric values
# It calculates y/x and returns the result

ARGS = {
    'ratio_string': '16:9'
}
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 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
exponential_decay_distribution Ran
ran on a synthesized input
SagiPolaczek/NeuralSVG/src/models/nerf_mlp_multi.py
code served (permissive licence) · get_code("b6bf865ea5dd217b")
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 ndarray, ndim 1, dtype family float, shape (5,) 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 numpy as np

ARGS = {
    'size': 5,
    'temperature': 1.5,
    'last_item_prob': 0.2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (5,) — shape (5,) 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
get_attention_map Ran
ran on a synthesized input
SagiPolaczek/NeuralSVG/src/models/attention_utils.py
code served (permissive licence) · get_code("9f3cdd853bc0c6f9")
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 ndarray, ndim 2, dtype family float, shape (224, 224) 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 a minimal mock model that has the structure expected by get_attention_map
class MockAttentionBlock(nn.Module):
    def __init__(self, num_tokens):
        super().__init__()
        self.attn_probs = torch.randn(1, 12, num_tokens, num_tokens)
    
    def forward(self, x):
        return x

class MockTransformer(nn.Module):
    def __init__(self, num_tokens):
        super().__init__()
        self.resblocks = nn.ModuleDict({
            'block_0': MockAttentionBlock(num_tokens),
            'block_1': MockAttentionBlock(num_tokens),
            'block_2': MockAttentionBlock(num_tokens),
        })
    
    def forward(self, x):
        return x

class MockVisual(nn.Module):
    def __init__(self, num_tokens):
        super().__init__()
        self.transformer = MockTransformer(num_tokens)
    
    def forward(self, x):
        return x

class MockCLIPModel(nn.Module):
    def __init__(self, num_tokens):
        super().__init__()
        self.visual = MockVisual(num_tokens)
    
    def encode_image(self, images, mode=None):
        # Simulate encoding by setting attn_probs on all blocks
        for block in self.visual.transformer.resblocks.values():
            block.attn_probs = torch.randn(1, 12, num_tokens, num_tokens)
        return images
    
    def zero_grad(self):
        super().zero_grad()

# Create the model with 50 tokens (7x7 patches + 1 class token)
num_tokens = 50
model = MockCLIPModel(num_tokens)

# Create a small image tensor (batch=1, channels=3, height=224, width=224)
image = torch.randn(1, 3, 224, 224)

# Device and image size
device = 'cpu'
image_size = 224

ARGS = {
    'image': image,
    'model': model,
    'device': device,
    'image_size': image_size,
}
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 ndarray, ndim 2, dtype family float, shape (224, 224) — shape (224, 224) 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
build_model Not yet run
unverified (harvested, no recorded run)
SagiPolaczek/NeuralSVG/CLIP_/clip/model.py
code served (permissive licence) · get_code("54a4f1da198cb291")
How it ran

no invocation recorded.

set_init_strokes_with_attention_map Not yet run
unverified (harvested, no recorded run)
SagiPolaczek/NeuralSVG/src/models/attention_utils.py
code served (permissive licence) · get_code("edd74ecca2518f03")
How it ran

no invocation recorded.

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 distinct functions with an identity, 1 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 1 the same month, 1 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
build_model 119 other papers
100 dated earlier, 1 the same month, 14 later, 2 same year, order not determinable, 2 undated
1809.11096 (2018-09)
Large Scale GAN Training for High Fidelity Natural Image Synthesis
1901.07441 (2019-01)
PadChest: A large chest x-ray image dataset with multi-label annotated reports
2104.14294 (2021-04)
Emerging Properties in Self-Supervised Vision Transformers
and 116 more not named here

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

The girl with a pearl earring" "A fountain pen" "A peacock" "A punk rocker" Figure 1. NeuralSVG generates vector graphics from text prompts with ordered and editable shapes. Our method supports dynamic background color conditioning, facilitating the generation of multiple color palettes for a single learned representation (right).

For agents

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

get_harvested_code_for_paper("2501.03992")
get_code_for_paper("2501.03992")
have("2501.03992")

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

curl https://syntology.ai/api/ran/2501.03992.json

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

[![Syntology run record](https://syntology.ai/api/ran/2501.03992.svg)](https://syntology.ai/paper/2501.03992)

Connect an agent — have() is free.