SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2206.10673 · 2022

Natural Backdoor Datasets

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/2206.10673.json · badge: /api/ran/2206.10673.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
uchicago-sandlab/naturalbackdoors canonical (classifier label)
classifier basis not recorded on the link
3 ran · 1 no recorded run
FunctionStatusWhere it lives
HMString Ran
ran on a synthesized input
uchicago-sandlab/naturalbackdoors/utils/gen_util.py
code served (permissive licence) · get_code("1f645e25b9ca237e")
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

torch.manual_seed(0)

# HMString takes a single parameter: sec_elapsed (seconds as a number)
# The function converts it to hours:minutes:seconds.milliseconds format
# It uses arithmetic operations on the parameter, so it needs a numeric value

sec_elapsed = 3661.5  # Example: 1 hour, 1 minute, 1.5 seconds

ARGS = {
    'sec_elapsed': sec_elapsed
}
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 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
assign_gpu Ran
ran on a synthesized input
uchicago-sandlab/naturalbackdoors/run_on_gpus_recreate_results.py
code served (permissive licence) · get_code("1de1a21276d533b2")
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 iterates over args (a list), checks if any element equals 'GPUID',
# and replaces it with str(gpu_idx). It then returns the modified args list.
# So args should be a list that may contain the string 'GPUID' and other elements.

args = ['model_path', 'GPUID', 'batch_size', 32, 'GPUID', 'output_dir']
gpu_idx = 2

ARGS = {
    'args': args,
    'gpu_idx': gpu_idx
}
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
timeit Ran
ran on a synthesized input
uchicago-sandlab/naturalbackdoors/utils/gen_util.py
code served (permissive licence) · get_code("c7283934eb2feadc")
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
import torch.nn as nn
import time

def timeit(method):
    def timed(*args, **kw):
        ts = time.time()
        result = method(*args, **kw)
        te = time.time()
        print('%r %f sec' % (method.__name__, te - ts))
        return result
    return timed

# Create a simple function to decorate
@timeit
def sample_function(x, y):
    """A simple function that the decorator will time."""
    time.sleep(0.001)
    return x + y

# Create a simple class with a method to decorate
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

torch.manual_seed(0)

# Build arguments for the timeit function
# timeit takes 'method' as its parameter
sample_method = lambda x, y: x + y

ARGS = {
    'method': sample_method
}

# Execute to verify it works
decorated_func = timeit(sample_method)
result = decorated_func(5, 3)
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
init_gpu Not yet run
unverified (harvested, no recorded run)
uchicago-sandlab/naturalbackdoors/utils/gen_util.py
code served (permissive licence) · get_code("dde0d5cd1e791ce4")
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

Extensive literature on backdoor poison attacks has studied attacks and defenses for backdoors using "digital trigger patterns." In contrast, "physical backdoors" use physical objects as triggers, have only recently been identified, and are qualitatively different enough to resist all defenses targeting digital trigger backdoors. Research on physical backdoors is limited by access to large datasets containing real images of physical objects co-located with targets of classification. Building these datasets is time- and labor-intensive. This works seeks to address the challenge of accessibility for research on physical backdoor attacks. We hypothesize that there may be naturally occurring physically co-located objects already present in popular datasets such as ImageNet. Once identified, a careful relabeling of these data can transform them into training samples for physical backdoor attacks. We propose a method to scalably identify these subsets of potential triggers in existing datasets, along with the specific classes they can poison. We call these naturally occurring trigger-class subsets natural backdoor datasets. Our techniques successfully identify natural backdoors in widely-available datasets, and produce models behaviorally equivalent to those trained on manually curated datasets. We release our code to allow the research community to create their own datasets for research on physical backdoor attacks.

For agents

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

get_harvested_code_for_paper("2206.10673")
get_code_for_paper("2206.10673")
have("2206.10673")

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

curl https://syntology.ai/api/ran/2206.10673.json

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

[![Syntology run record](https://syntology.ai/api/ran/2206.10673.svg)](https://syntology.ai/paper/2206.10673)

Connect an agent — have() is free.