SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1908.11515 · 2019

Improving Utility and Security of the Shuffler-based Differential Privacy

arXiv · PDF · Open in the Atlas

Code that ran

8 of 13 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 8 have no contract verdict; 0 failures of our instrument; 5 no recorded run), measured 2026-11-21; 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/1908.11515.json · badge: /api/ran/1908.11515.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
vvv214/LDP_Protocols canonical (classifier label)
classifier basis: census
8 ran · 5 no recorded run
FunctionStatusWhere it lives
EM Ran
ran on a synthesized input
vvv214/LDP_Protocols/sw.py
code served (permissive licence) · get_code("a132c0ee784427bc")
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

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

# Define parameters for the EM function
n = 4  # number of components/states
ns_hist = np.array([10, 15, 8, 12])  # histogram of observations (sample counts)
transform = np.array([
    [0.8, 0.1, 0.05, 0.05],
    [0.1, 0.7, 0.1, 0.1],
    [0.05, 0.15, 0.7, 0.1],
    [0.05, 0.05, 0.15, 0.75]
])  # transformation matrix (4x4)
max_iteration = 100  # maximum number of iterations
loglikelihood_threshold = 1e-4  # convergence threshold for log-likelihood improvement

ARGS = {
    "n": n,
    "ns_hist": ns_hist,
    "transform": transform,
    "max_iteration": max_iteration,
    "loglikelihood_threshold": loglikelihood_threshold
}
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
blanket Ran
ran on a synthesized input
vvv214/LDP_Protocols/murs/analysis.py
code served (permissive licence) · get_code("4b1e85aed933904e")
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 math

# The function takes two parameters: coeff and binom
# Both are used in a division and then passed to math.sqrt
# They should be numeric values (int or float)

ARGS = {
    "coeff": 16.0,
    "binom": 4.0,
}
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 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
blanket_naive Ran
ran on a synthesized input
vvv214/LDP_Protocols/murs/analysis.py
code served (permissive licence) · get_code("cd0853e7e5bfd09d")
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 math

# Analyze the function blanket_naive:
# - eps0: float, used in math.exp(eps0)
# - delta: float, used in math.log(2 / delta)
# - n: int or float, used in arithmetic (n - 1)
# - d: int or float, used in arithmetic (d - 1)
# All parameters are scalars used in mathematical operations

ARGS = {
    "eps0": 0.5,
    "delta": 0.1,
    "n": 100,
    "d": 10
}
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 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
blanket_naive_olh Ran
ran on a synthesized input
vvv214/LDP_Protocols/murs/analysis.py
code served (permissive licence) · get_code("c2b62391ab3e2b9e")
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 math

# Analyze the function blanket_naive_olh(eps0, delta, n)
# Parameters:
# - eps0: used in math.exp(eps0 / 2) and math.exp(eps0), so a float
# - delta: used in math.log(2 / delta), so a positive float
# - n: used in (n - 1) division, so a number (int or float)

ARGS = {
    "eps0": 0.5,
    "delta": 0.1,
    "n": 100
}
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 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
build_range_tree Ran
ran on a synthesized input
vvv214/LDP_Protocols/hio/range_tree.py
code served (permissive licence) · get_code("c83528b6198f50d9")
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 Tree. 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.

class DummyClass:
    def build_range_tree(self, l, r, fanout, cur_layer, total_layer):
        tree = Tree()
        tree.range = (l, r)
        if cur_layer == total_layer:
            return tree
        step_size = (r - l) / fanout
        cur_l = l
        for i in range(fanout):
            cur_r = l + int(step_size * (i + 1))
            if not cur_r == cur_l:
                tree.children.append(self.build_range_tree(cur_l, cur_r, fanout, cur_layer + 1, total_layer))
                cur_l = cur_r
        return tree

class Tree(object):
    def __init__(self):
        self.children = []
        self.range = None

dummy_instance = DummyClass()

ARGS = {
    "self": dummy_instance,
    "l": 0,
    "r": 100,
    "fanout": 2,
    "cur_layer": 1,
    "total_layer": 3
}
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 Tree — a Tree output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Tree 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
elgamal_dec Ran
ran on a synthesized input
vvv214/LDP_Protocols/murs/ec_crypto.py
code served (permissive licence) · get_code("88e753c593c5eb8c")
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 EGPoint. 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)

# The function uses:
# - c1 * priv_key (multiplication)
# - (c1 * priv_key).x (accessing .x attribute)
# - c2 - ... (subtraction)
# So we need objects that support these operations

class EGPoint:
    """Represents an elliptic curve point with x coordinate"""
    def __init__(self, x, y=None):
        self.x = x
        self.y = y
    
    def __mul__(self, scalar):
        """Scalar multiplication on elliptic curve"""
        # Returns another EGPoint
        return EGPoint(self.x * scalar, self.y * scalar if self.y is not None else None)
    
    def __rmul__(self, scalar):
        """Right multiplication"""
        return self.__mul__(scalar)
    
    def __sub__(self, other):
        """Point subtraction"""
        if isinstance(other, EGPoint):
            return EGPoint(self.x - other.x, self.y - other.y if self.y is not None else None)
        return EGPoint(self.x - other, self.y)

# Create test data
c1 = EGPoint(torch.tensor([1.0, 2.0, 3.0, 4.0]), torch.tensor([5.0, 6.0, 7.0, 8.0]))
c2 = EGPoint(torch.tensor([10.0, 20.0, 30.0, 40.0]), torch.tensor([50.0, 60.0, 70.0, 80.0]))
priv_key = torch.tensor([2.0, 3.0, 4.0, 5.0])

ARGS = {
    'c1': c1,
    'c2': c2,
    'priv_key': priv_key
}
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 EGPoint — a EGPoint output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a EGPoint 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
sw Ran
ran on a synthesized input
vvv214/LDP_Protocols/sw.py
code served (permissive licence) · get_code("b2d9f45140070391")
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 (8,) 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(0)

# Create small sample data for testing
ori_samples = np.array([0.2, 0.3, 0.5, 0.1, 0.4, 0.6, 0.25, 0.75])

# Define parameters
l = 0.0  # lower bound
h = 1.0  # upper bound
eps = 0.5  # privacy budget
randomized_bins = 16  # number of randomized bins (smaller for testing)
domain_bins = 8  # number of domain bins (smaller for testing)

ARGS = {
    "ori_samples": ori_samples,
    "l": l,
    "h": h,
    "eps": eps,
    "randomized_bins": randomized_bins,
    "domain_bins": domain_bins,
}
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 (8,) — shape (8,) 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
traverse_range_tree Ran
ran on a synthesized input
vvv214/LDP_Protocols/hio/range_tree.py
code served (permissive licence) · get_code("f8a73171cb0f9bab")
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 torch
import numpy as np

class RangeNode:
    def __init__(self, range_val, children=None):
        self.range = range_val
        self.children = children if children is not None else []

# Create a simple tree structure
# Root node
root = RangeNode(
    range_val=(0, 10),
    children=[
        RangeNode(
            range_val=(0, 5),
            children=[
                RangeNode(range_val=(0, 2), children=[]),
                RangeNode(range_val=(3, 5), children=[])
            ]
        ),
        RangeNode(
            range_val=(6, 10),
            children=[
                RangeNode(range_val=(6, 8), children=[]),
                RangeNode(range_val=(9, 10), children=[])
            ]
        )
    ]
)

ARGS = {
    "root": root
}
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 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
EMS Not yet run
unverified (harvested, no recorded run)
vvv214/LDP_Protocols/sw.py
code served (permissive licence) · get_code("3cd53663eafdfcb3")
How it ran

no invocation recorded.

evaluate Not yet run
unverified (harvested, no recorded run)
vvv214/LDP_Protocols/pem/exp.py
code served (permissive licence) · get_code("deb84a6ea4773da4")
How it ran

no invocation recorded.

lh Not yet run
unverified (harvested, no recorded run)
vvv214/LDP_Protocols/pem/fo.py
code served (permissive licence) · get_code("ef3081e80fa04f63")
How it ran

no invocation recorded.

lh_aggregate Not yet run
unverified (harvested, no recorded run)
vvv214/LDP_Protocols/pem/fo.py
code served (permissive licence) · get_code("adec0e0a0b645dfa")
How it ran

no invocation recorded.

lh_perturb Not yet run
unverified (harvested, no recorded run)
vvv214/LDP_Protocols/pem/fo.py
code served (permissive licence) · get_code("7af17629cdb65c23")
How it ran

no invocation recorded.

Also present in other papers' code

Of 13 samples lifted from this paper's linked repositories, 13 carry a function identity and 0 do not. Of the 13 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 13 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 2026-11-21.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

When collecting information, local differential privacy (LDP) alleviates privacy concerns of users because their private information is randomized before being sent it to the central aggregator. LDP imposes large amount of noise as each user executes the randomization independently. To address this issue, recent work introduced an intermediate server with the assumption that this intermediate server does not collude with the aggregator. Under this assumption, less noise can be added to achieve the same privacy guarantee as LDP, thus improving utility for the data collection task. This paper investigates this multiple-party setting of LDP. We analyze the system model and identify potential adversaries. We then make two improvements: a new algorithm that achieves a better privacy-utility tradeoff; and a novel protocol that provides better protection against various attacks. Finally, we perform experiments to compare different methods and demonstrate the benefits of using our proposed method.

For agents

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

get_harvested_code_for_paper("1908.11515")
get_code_for_paper("1908.11515")
have("1908.11515")

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

curl https://syntology.ai/api/ran/1908.11515.json

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

[![Syntology run record](https://syntology.ai/api/ran/1908.11515.svg)](https://syntology.ai/paper/1908.11515)

Connect an agent — have() is free.