SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2310.20673 · ICLR · 2024

Balancing Act: Constraining Disparate Impact in Sparse Models

Simon Lacoste-Julien, Rohan Sukumaran, Jose Gallego-Posada, Juan Ramirez, Meraj Hashemizadeh, Golnoosh Farnadi

arXiv · PDF · Open in the Atlas

Code that ran

2 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; 1 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 1 failure 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/2310.20673.json · badge: /api/ran/2310.20673.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
merajhashemi/balancing-act canonical (classifier label)
classifier basis: census
2 ran · 2 no recorded run
FunctionStatusWhere it lives
get_num_classes_and_protected_groups Ran
ran, honoured the contract we drafted
merajhashemi/balancing-act/src/datasets/datasets.py
code served (permissive licence) · get_code("7dca39b003cc513f")
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 int; 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.utils.data as torch_data

torch.manual_seed(0)

class StandInDataset:
    def __init__(self, num_classes, num_protected_groups):
        self.num_classes = num_classes
        self.num_protected_groups = num_protected_groups

class SubsetDataset(torch_data.Subset):
    def __init__(self, dataset, indices):
        self.dataset = dataset
        self.indices = indices

# Create a base dataset with num_classes and num_protected_groups attributes
base_dataset = StandInDataset(num_classes=10, num_protected_groups=2)

# Test case 1: Direct dataset (not a Subset)
dataset_direct = base_dataset

# Test case 2: Subset wrapping a dataset
subset_dataset = SubsetDataset(base_dataset, indices=[0, 1, 2])

ARGS = {
    "dataset": dataset_direct
}
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 tuple, ndim 1, dtype family int — 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
unpack_indices_from_protected_attr Ran
ran; our contract draft was wrong, not the code
merajhashemi/balancing-act/src/cmp/utils.py
code served (permissive licence) · get_code("069d0a228ee418cd")
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 int, shape (1, 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 numpy as np

torch.manual_seed(0)

# Test case 1: non-intersectional with multiple protected attributes
protected_attr_1 = torch.tensor([
    [0, 1],
    [1, 0]
], dtype=torch.long)

num_protected_groups_1 = [3, 4]
intersectional_1 = False

ARGS = {
    "protected_attr": protected_attr_1,
    "num_protected_groups": num_protected_groups_1,
    "intersectional": intersectional_1
}
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 Tensor, ndim 2, dtype family int, shape (1, 4) — shape (1, 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
add_dense_to_sparse Not yet run
unverified (harvested, no recorded run)
merajhashemi/balancing-act/src/cmp/utils.py
code served (permissive licence) · get_code("9167b727d3da64c0")
How it ran

no invocation recorded.

unpack_sparse_and_pad Not yet run
unverified (harvested, no recorded run)
merajhashemi/balancing-act/src/cmp/utils.py
code served (permissive licence) · get_code("1ede60e119687f79")
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

Model pruning is a popular approach to enable the deployment of large deep learning models on edge devices with restricted computational or storage capacities. Although sparse models achieve performance comparable to that of their dense counterparts at the level of the entire dataset, they exhibit high accuracy drops for some data sub-groups. Existing methods to mitigate this disparate impact induced by pruning (i) rely on surrogate metrics that address the problem indirectly and have limited interpretability; or (ii) scale poorly with the number of protected sub-groups in terms of computational cost. We propose a constrained optimization approach that directly addresses the disparate impact of pruning: our formulation bounds the accuracy change between the dense and sparse models, for each subgroup. This choice of constraints provides an interpretable success criterion to determine if a pruned model achieves acceptable disparity levels. Experimental results demonstrate that our technique scales reliably to problems involving large models and hundreds of protected sub-groups.

For agents

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

get_harvested_code_for_paper("2310.20673")
get_code_for_paper("2310.20673")
have("2310.20673")

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

curl https://syntology.ai/api/ran/2310.20673.json

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

[![Syntology run record](https://syntology.ai/api/ran/2310.20673.svg)](https://syntology.ai/paper/2310.20673)

Connect an agent — have() is free.