SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1806.00035 · 2018

Assessing Generative Models via Precision and Recall

arXiv · PDF · Open in the Atlas

Code that ran

10 of 15 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 6 have no contract verdict; 4 failures of our instrument; 5 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/1806.00035.json · badge: /api/ran/1806.00035.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
msmsajjadi/precision-recall-distributions canonical (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
ok1zjf/lbae community (classifier label)
classifier basis not recorded on the link
7 ran · 5 no recorded run
FunctionStatusWhere it lives
calculate_frechet_distance Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ok1zjf/lbae/fid.py
code served (permissive licence) · get_code("5277c96076ea2fe5")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned float64, 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 numpy as np

np.random.seed(42)

# Create mean vectors for two distributions
# Using different dimensions to make them distinct
feature_dim = 128

mu1 = np.random.randn(feature_dim)
mu2 = np.random.randn(feature_dim)

# Create covariance matrices (must be positive semi-definite)
# Generate random covariance matrices by A @ A.T method
A1 = np.random.randn(feature_dim, feature_dim)
sigma1 = A1 @ A1.T

A2 = np.random.randn(feature_dim, feature_dim)
sigma2 = A2 @ A2.T

eps = 1e-06

ARGS = {
    'mu1': mu1,
    'sigma1': sigma1,
    'mu2': mu2,
    'sigma2': sigma2,
    'eps': eps
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, 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
compute_prd Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
msmsajjadi/precision-recall-distributions/prd_score.py
code served (permissive licence) · get_code("b9da61c36d3b8f11")
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) and returned tuple, ndim 2, 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

# Create valid probability distributions
# eval_dist: probabilities for the distribution to be evaluated
eval_dist = np.array([0.1, 0.2, 0.3, 0.4])

# ref_dist: probabilities for the reference distribution
ref_dist = np.array([0.25, 0.25, 0.25, 0.25])

# num_angles: number of angles for PRD computation (must be in [3, 1e6])
num_angles = 1001

# epsilon: angle for edge cases (must be in (0, 0.1])
epsilon = 1e-10

ARGS = {
    'eval_dist': eval_dist,
    'ref_dist': ref_dist,
    'num_angles': num_angles,
    'epsilon': epsilon,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 2, 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
compute_prd_from_embedding Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
msmsajjadi/precision-recall-distributions/prd_score.py
code served (permissive licence) · get_code("5d84440e02c3a502")
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) and returned tuple, ndim 2, 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 sklearn.cluster

# Set random seed for reproducibility
np.random.seed(0)

# Create small sample embeddings for eval and ref distributions
# Using distinct dimensions: 8 samples, 4-dimensional embeddings
eval_data = np.random.randn(8, 4).astype(np.float32)
ref_data = np.random.randn(8, 4).astype(np.float32)

# Define arguments for compute_prd_from_embedding
ARGS = {
    'eval_data': eval_data,
    'ref_data': ref_data,
    'num_clusters': 5,
    'num_angles': 101,
    'num_runs': 2,
    'enforce_balance': True,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, sklearn — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 2, 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
corrupt Ran
ran on a synthesized input
ok1zjf/lbae/datasets.py
code served (permissive licence) · get_code("0bee687e2a3cbee8")
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 Tensor, ndim 2, dtype family float, shape (2, 4) 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 small tensor for x
x = torch.randn(2, 4)

# Since the function body is just "return x", it doesn't actually use
# corrupt_method or corrupt_args for any specific operations.
# We provide minimal valid arguments.
corrupt_method = "none"
corrupt_args = {}

ARGS = {
    "x": x,
    "corrupt_method": corrupt_method,
    "corrupt_args": corrupt_args,
}
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 Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) 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
norm Ran
ran on a synthesized input
ok1zjf/lbae/sampler.py
code served (permissive licence) · get_code("13d8e00ecc2023db")
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 (4,) 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

torch_tensor = np.array([1.0, 2.0, 3.0, 4.0])

ARGS = {
    "a": torch_tensor
}
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 (4,) — shape (4,) 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
prd_to_max_f_beta_pair Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
msmsajjadi/precision-recall-distributions/prd_score.py
code served (permissive licence) · get_code("5adbefd0ebfa3e5e")
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) 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

# Create valid precision and recall arrays
precision = np.array([0.0, 0.25, 0.5, 0.75, 1.0])
recall = np.array([0.1, 0.3, 0.6, 0.8, 0.95])

# Create ARGS dict with all parameters for prd_to_max_f_beta_pair
ARGS = {
    'precision': precision,
    'recall': recall,
    'beta': 8
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
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
psnr Ran
ran on a synthesized input
ok1zjf/lbae/image_utils.py
code served (permissive licence) · get_code("8cdf564a25c20149")
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 float64, ndim 0, 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 torch
torch.manual_seed(0)

# The function psnr takes two parameters: original and reconstruction
# Both are used with peak_signal_noise_ratio which expects numpy arrays
# The function also iterates over them with zip and accesses .shape[0]
# So they should be numpy arrays (or array-like objects)

# Create small sample images as numpy arrays
# Using uint8 range [0, 255] as typical for image data
original = np.random.randint(0, 256, size=(2, 4, 3), dtype=np.uint8)
reconstruction = np.random.randint(0, 256, size=(2, 4, 3), dtype=np.uint8)

ARGS = {
    "original": original,
    "reconstruction": reconstruction
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, skimage, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
run_command Ran
ran on a synthesized input
ok1zjf/lbae/sys_utils.py
code served (permissive licence) · get_code("1bd4c31f00d84b70")
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 subprocess

def run_command(command):
    p = subprocess.Popen(command.split(), stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
    return '\n'.join(['\t' + line.decode('utf-8').strip() for line in p.stdout.readlines()])

ARGS = {
    "command": "echo hello world"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — 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
todeg Ran
ran on a synthesized input
ok1zjf/lbae/sampler.py
code served (permissive licence) · get_code("0047ef4dfbce17cc")
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 numpy as np
import torch

torch.manual_seed(0)

# The function todeg takes a parameter x and applies np.arccos to it,
# then converts to degrees. The input should be a value or array 
# with values in the range [-1, 1] (valid domain for arccos).

x = np.array([0.0, 0.5, 1.0, -1.0])

ARGS = {
    'x': x
}
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
tohms Ran
ran on a synthesized input
ok1zjf/lbae/sys_utils.py
code served (permissive licence) · get_code("c232fb314605487f")
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

def tohms(s):
    m, s = divmod(s, 60)
    h, m = divmod(m, 60)
    d, h = divmod(h, 24)
    time_str = '{:>02d}:{:>02}:{:>02d}'.format(int(h), int(m), int(s))
    return time_str

# The function takes a single parameter 's' which is a number representing seconds
# It performs arithmetic operations on it (divmod) and formats it as a time string
# We need to provide a numeric value (int or float) for 's'

ARGS = {
    's': 3661  # 1 hour, 1 minute, 1 second
}
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
calculate_activation_statistics Not yet run
unverified (harvested, no recorded run)
ok1zjf/lbae/fid.py
code served (permissive licence) · get_code("664d8c6c98c4ce7c")
How it ran

no invocation recorded.

get_activations Not yet run
unverified (harvested, no recorded run)
ok1zjf/lbae/fid.py
code served (permissive licence) · get_code("579daa09f4887be8")
How it ran

no invocation recorded.

nearestPDX Not yet run
unverified (harvested, no recorded run)
ok1zjf/lbae/sampler.py
code served (permissive licence) · get_code("5a075ff84473359d")
How it ran

no invocation recorded.

save_image Not yet run
unverified (harvested, no recorded run)
ok1zjf/lbae/image_utils.py
code served (permissive licence) · get_code("afc8d137b43810a4")
How it ran

no invocation recorded.

ssim Not yet run
unverified (harvested, no recorded run)
ok1zjf/lbae/image_utils.py
code served (permissive licence) · get_code("239d3d651c2bea69")
How it ran

no invocation recorded.

Also present in other papers' code

Of 15 samples lifted from this paper's linked repositories, 15 carry a function identity and 0 do not. Of the 15 distinct functions with an identity, 4 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 4 later) and 11 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
calculate_frechet_distance 15 other papers
4 dated earlier, 11 later
1406.2661 (2014-06)
Generative Adversarial Networks
1703.10593 (2017-03)
Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks
1706.08500 (2017-06)
GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
and 12 more not named here
compute_prd 3 other papers
3 later
1905.10485 (2019-05)
Generative Latent Flow
1912.04958 (2019-12)
Analyzing and Improving the Image Quality of StyleGAN
2010.15639 (2020-10)
Teaching a GAN What Not to Learn
compute_prd_from_embedding 3 other papers
3 later
1807.04720 (2018-07)
A Large-Scale Study on Regularization and Normalization in GANs
1905.10485 (2019-05)
Generative Latent Flow
2010.15639 (2020-10)
Teaching a GAN What Not to Learn
prd_to_max_f_beta_pair 3 other papers
3 later
1807.04720 (2018-07)
A Large-Scale Study on Regularization and Normalization in GANs
1905.10485 (2019-05)
Generative Latent Flow
2010.15639 (2020-10)
Teaching a GAN What Not to Learn

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

Recent advances in generative modeling have led to an increased interest in the study of statistical divergences as means of model comparison. Commonly used evaluation methods, such as the Frechet Inception Distance (FID), correlate well with the perceived quality of samples and are sensitive to mode dropping. However, these metrics are unable to distinguish between different failure cases since they only yield one-dimensional scores. We propose a novel definition of precision and recall for distributions which disentangles the divergence into two separate dimensions. The proposed notion is intuitive, retains desirable properties, and naturally leads to an efficient algorithm that can be used to evaluate generative models. We relate this notion to total variation as well as to recent evaluation metrics such as Inception Score and FID. To demonstrate the practical utility of the proposed approach we perform an empirical study on several variants of Generative Adversarial Networks and Variational Autoencoders. In an extensive set of experiments we show that the proposed metric is able to disentangle the quality of generated samples from the coverage of the target distribution.

For agents

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

get_harvested_code_for_paper("1806.00035")
get_code_for_paper("1806.00035")
have("1806.00035")

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

curl https://syntology.ai/api/ran/1806.00035.json

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

[![Syntology run record](https://syntology.ai/api/ran/1806.00035.svg)](https://syntology.ai/paper/1806.00035)

Connect an agent — have() is free.