SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2004.03657 · 2020

FedMAX: Mitigating Activation Divergence for Accurate and Communication-Efficient Federated Learning

arXiv · PDF · Open in the Atlas

Code that ran

12 of 14 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; 1 honoured the contract we drafted, 1 violated it and 9 have no contract verdict; 1 failure of our instrument; 2 no recorded run); 1 of them was also run on the host, not in the jail, by the peer-agreement lane that set its level, 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/2004.03657.json · badge: /api/ran/2004.03657.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
weichennone/FedMAX community (classifier label)
classifier basis not recorded on the link
12 ran · 2 no recorded run
FunctionStatusWhere it lives
count_params Ran
ran, honoured the contract we drafted
weichennone/FedMAX/medical_data/lib/utils.py
pointer only (licence: MIT); the stored copy get_code serves, from gdisag/gradient_disaggregation, is pointer only (licence: NONE) · get_code("616b5e82dd2a38f9")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from gdisag/gradient_disaggregation, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

FedAvg Ran
ran; our contract draft was wrong, not the code
weichennone/FedMAX/digit_object_recognition/models/Fed.py
pointer only (licence: MIT); the stored copy get_code serves, from hongshenghu/source-inference-fl, is pointer only (licence: NONE) · get_code("f9964d719c5cb8f4")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from hongshenghu/source-inference-fl, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

cifar_iid Ran
ran on a synthesized input
weichennone/FedMAX/digit_object_recognition/utils/sampling.py
code served (permissive licence) · get_code("6b0f40ddd37b7b2c")
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 dict. 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
from torch.utils.data import Dataset

class DummyCIFAR10(Dataset):
    """Minimal stand-in for CIFAR10 dataset"""
    def __init__(self, size=100):
        self.size = size
        self.data = np.random.randint(0, 256, (size, 32, 32, 3), dtype=np.uint8)
        self.targets = np.random.randint(0, 10, size)
    
    def __len__(self):
        return self.size
    
    def __getitem__(self, idx):
        return self.data[idx], self.targets[idx]

np.random.seed(42)
dummy_dataset = DummyCIFAR10(size=100)
num_users = 10

ARGS = {
    'dataset': dummy_dataset,
    'num_users': num_users,
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
cifar_noniid_2 Ran
ran on a synthesized input
weichennone/FedMAX/digit_object_recognition/utils/sampling.py
code served (permissive licence) · get_code("a2be42588b41c16d")
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 dict. 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

class MockCIFAR10Dataset:
    """Mock CIFAR-10 dataset with targets attribute"""
    def __init__(self, num_samples=50000):
        self.num_samples = num_samples
        # Create mock targets - 10 classes, distributed across samples
        self.targets = np.random.randint(0, 10, num_samples)
    
    def __len__(self):
        return self.num_samples

# Create mock dataset
dataset = MockCIFAR10Dataset(num_samples=50000)

# Number of users/clients
num_users = 10

ARGS = {
    "dataset": dataset,
    "num_users": num_users,
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
confusion_mat Ran
ran on a synthesized input
weichennone/FedMAX/medical_data/lib/metrics.py
code served (permissive licence) · get_code("28952daced30ce4d")
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 int, shape (3, 3) 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
from sklearn import metrics

torch.manual_seed(0)

# y_pred can be either:
# - a 2D tensor/array with shape (batch, num_classes) -> argmax is taken
# - a 2D tensor/array with shape (batch, 1) -> squeezed to 1D
# - a 1D tensor/array -> used as-is
# Let's create a case with shape (batch, num_classes) where num_classes > 1

y_pred_tensor = torch.randn(2, 4)  # batch=2, num_classes=4
y_true_tensor = torch.tensor([0, 2])  # batch=2, true labels

ARGS = {
    "y_pred": y_pred_tensor,
    "y_true": y_true_tensor,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn, 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 int, shape (3, 3) — shape (3, 3) 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
f1_micro Ran
ran on a synthesized input
weichennone/FedMAX/medical_data/lib/metrics.py
code served (permissive licence) · get_code("5c35afeaa0ba402f")
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

torch.manual_seed(0)

# Create y_pred as a tensor with shape (2, 4) - batch size 2, 4 classes
# Using multi-class format (not single column)
y_pred = torch.randn(2, 4)

# Create y_true as a tensor with shape (2,) - batch size 2, true class labels
y_true = torch.tensor([0, 2])

ARGS = {
    "y_pred": y_pred,
    "y_true": y_true,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn, 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
fedavg Ran
ran on a synthesized input
weichennone/FedMAX/medical_data/fed.py
code served (permissive licence) · get_code("55fcb483356a9173")
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 dict. 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 copy
import torch
import torch.nn as nn

torch.manual_seed(0)

# The function fedavg takes a list of weight dictionaries and averages them
# Each weight dictionary should have tensor values

# Create sample weight dictionaries (simulating model state_dicts)
w1 = {
    'layer1.weight': torch.randn(4, 3),
    'layer1.bias': torch.randn(4),
    'layer2.weight': torch.randn(5, 4),
}

w2 = {
    'layer1.weight': torch.randn(4, 3),
    'layer1.bias': torch.randn(4),
    'layer2.weight': torch.randn(5, 4),
}

w3 = {
    'layer1.weight': torch.randn(4, 3),
    'layer1.bias': torch.randn(4),
    'layer2.weight': torch.randn(5, 4),
}

# The function expects a list of weight dicts
w = [w1, w2, w3]

ARGS = {
    'w': w
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
normalize Ran
ran on a synthesized input
weichennone/FedMAX/medical_data/lib/preprocess.py
code served (permissive licence) · get_code("f7cae9b5a404a4a3")
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 int, shape (4, 3) 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 cv2
import numpy as np

# Create a small test image
# The function uses cv2.addWeighted and cv2.GaussianBlur which expect numpy arrays
# img_size is used in the calculation img_size / 30 for the Gaussian blur sigma
src = np.random.randint(0, 256, (4, 3), dtype=np.uint8)
img_size = 30

ARGS = {
    "src": src,
    "img_size": img_size
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy — 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 int, shape (4, 3) — shape (4, 3) 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
quadratic_weighted_kappa Ran
ran on a synthesized input
weichennone/FedMAX/medical_data/lib/metrics.py
code served (permissive licence) · get_code("32a4630f4052c270")
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
from sklearn import metrics

torch.manual_seed(0)

# Case 1: y_pred is a 2D tensor with multiple columns (will use argmax)
y_pred_tensor = torch.randn(2, 4)  # batch size 2, 4 classes
y_true_tensor = torch.tensor([0, 2], dtype=torch.long)  # batch size 2, class labels

ARGS = {
    "y_pred": y_pred_tensor,
    "y_true": y_true_tensor,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn, 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
remove_boundaries Ran
ran on a synthesized input
weichennone/FedMAX/medical_data/lib/preprocess.py
code served (permissive licence) · get_code("6e2803e2d089d093")
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 3, dtype family float, shape (64, 64, 3) 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
import cv2

# Create a small test image - 3D array with shape (height, width, channels)
# Using shape (64, 64, 3) for RGB image
src = np.random.randint(0, 256, size=(64, 64, 3), dtype=np.uint8).astype(np.float32)

# img_size parameter - controls the radius of the circular mask
img_size = 64

ARGS = {
    "src": src,
    "img_size": img_size
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family float, shape (64, 64, 3) — shape (64, 64, 3) 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
scale_radius Ran
ran on a synthesized input
weichennone/FedMAX/medical_data/lib/preprocess.py
code served (permissive licence) · get_code("0079a1a6108a54b6")
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 3, dtype family int, shape (32, 32, 3) 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
import cv2
from skimage import measure

# Create a small test image with 3 channels (RGB-like)
# Shape: (height, width, channels) = (64, 64, 3)
src_image = np.random.randint(0, 256, size=(64, 64, 3), dtype=np.uint8)

# Ensure the image has some structure so regionprops can find a region
# Create a bright circular region in the center
y, x = np.ogrid[:64, :64]
center = (32, 32)
radius = 15
mask = (x - center[1])**2 + (y - center[0])**2 <= radius**2
src_image[mask] = 200

ARGS = {
    "src": src_image,
    "img_size": 32,
    "padding": False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy, skimage — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family int, shape (32, 32, 3) — shape (32, 32, 3) 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
str2bool Ran
ran, violated the contract we drafted
weichennone/FedMAX/medical_data/lib/utils.py
code served (permissive licence) · get_code("2e468cbd0d4a4241")
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 bool, 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 argparse

def str2bool(v):
    if v.lower() in ['true', 1]:
        return True
    elif v.lower() in ['false', 0]:
        return False
    else:
        raise argparse.ArgumentTypeError('Boolean value expected.')

ARGS = {
    'v': 'true'
}
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 bool, 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
accuracy Not yet run
unverified (harvested, no recorded run)
weichennone/FedMAX/medical_data/lib/utils.py
pointer only (licence: MIT); the stored copy get_code serves, from boone891214/MEST, is pointer only (licence: NONE) · get_code("913d82af53065418")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from boone891214/MEST, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

femnist_star Not yet run
unverified (harvested, no recorded run)
weichennone/FedMAX/digit_object_recognition/utils/sampling.py
code served (permissive licence) · get_code("f08f5ccff4b9f848")
How it ran

no invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 distinct functions with an identity, 4 are also present in other papers whose code we hold (4 with at least one paper dated earlier, 2 the same month, 4 later) and 10 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
accuracy 80 other papers
31 dated earlier, 4 the same month, 45 later
1307.0414 (2013-07)
Challenges in Representation Learning: A report on three machine learning contests
1412.6553 (2014-12)
Speeding-up Convolutional Neural Networks Using Fine-tuned CP-Decomposition
1503.02531 (2015-03)
Distilling the Knowledge in a Neural Network
and 77 more not named here
count_params 13 other papers
3 dated earlier, 10 later
1807.10165 (2018-07)
UNet++: A Nested U-Net Architecture for Medical Image Segmentation
1811.11168 (2018-11)
Deformable ConvNets v2: More Deformable, Better Results
1907.03950 (2019-07)
Learning by Abstraction: The Neural State Machine
and 10 more not named here
FedAvg 9 other papers
4 dated earlier, 1 the same month, 4 later
1602.05629 (2016-02)
Communication-Efficient Learning of Deep Networks from Decentralized Data
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
1812.07108 (2018-12)
Learning Private Neural Language Modeling with Attentive Aggregation
and 6 more not named here
str2bool 5 other papers
2 dated earlier, 3 later
1807.10165 (2018-07)
UNet++: A Nested U-Net Architecture for Medical Image Segmentation
1811.11168 (2018-11)
Deformable ConvNets v2: More Deformable, Better Results
2106.00487 (2021-06)
Dense Nested Attention Network for Infrared Small Target Detection
and 2 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

In this paper, we identify a new phenomenon called activation-divergence which occurs in Federated Learning (FL) due to data heterogeneity (i.e., data being non-IID) across multiple users. Specifically, we argue that the activation vectors in FL can diverge, even if subsets of users share a few common classes with data residing on different devices. To address the activation-divergence issue, we introduce a prior based on the principle of maximum entropy; this prior assumes minimal information about the per-device activation vectors and aims at making the activation vectors of same classes as similar as possible across multiple devices. Our results show that, for both IID and non-IID settings, our proposed approach results in better accuracy (due to the significantly more similar activation vectors across multiple devices), and is more communication-efficient than state-of-the-art approaches in FL. Finally, we illustrate the effectiveness of our approach on a few common benchmarks and two large medical datasets.

For agents

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

get_harvested_code_for_paper("2004.03657")
get_code_for_paper("2004.03657")
have("2004.03657")

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

curl https://syntology.ai/api/ran/2004.03657.json

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

[![Syntology run record](https://syntology.ai/api/ran/2004.03657.svg)](https://syntology.ai/paper/2004.03657)

Connect an agent — have() is free.