SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2410.04472 · ICLR · 2025

Collapsed Language Models Promote Fairness

Linyi Li, Wuyang Chen, Yunchao Wei, Yao Zhao, Jingxuan Xu

arXiv · PDF · Open in the Atlas

Code that ran

12 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 0 have no contract verdict; 12 failures of our instrument; 1 no recorded run), measured 2026-11-22; 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/2410.04472.json · badge: /api/ran/2410.04472.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
Xujxyang/Fairness-NC-main canonical (classifier label)
classifier basis: census
11 ran · 1 no recorded run
xujxyang/fairness-nc-main canonical (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
FunctionStatusWhere it lives
calculate_group_to_group_relative_distance_asymmetric Ran
ran; our contract draft was wrong, not the code
Xujxyang/Fairness-NC-main/adept/distance.py
pointer only (licence: NONE) · get_code("31501071e57d43c4")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Xujxyang/Fairness-NC-main 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.

calculate_group_to_group_relative_distance_asymmetric_test Ran
ran; our contract draft was wrong, not the code
Xujxyang/Fairness-NC-main/adept/distance.py
pointer only (licence: NONE) · get_code("d6b5586516edd277")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Xujxyang/Fairness-NC-main 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.

calculate_group_to_one_relative_distance_asymmetric_test Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Xujxyang/Fairness-NC-main/adept/distance.py
pointer only (licence: NONE) · get_code("d73ab432982f8191")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Xujxyang/Fairness-NC-main 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.

create_dataset Ran
ran; our contract draft was wrong, not the code
Xujxyang/Fairness-NC-main/adept/debias.py
pointer only (licence: NONE) · get_code("5f516b6aaad9acb6")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Xujxyang/Fairness-NC-main 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.

create_nc_dataset Ran
ran; our contract draft was wrong, not the code
Xujxyang/Fairness-NC-main/adept/debias.py
pointer only (licence: NONE) · get_code("f10b5f9242a0af99")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Xujxyang/Fairness-NC-main 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.

load_and_cache_examples Ran
ran; our contract draft was wrong, not the code
Xujxyang/Fairness-NC-main/adept/debias_original.py
pointer only (licence: NONE) · get_code("f46fa5e6d0948125")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Xujxyang/Fairness-NC-main 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.

process_batch Ran
ran; our contract draft was wrong, not the code
xujxyang/fairness-nc-main/bec/code/main_nc.py
pointer only (licence: NONE) · get_code("fe1027358a8d7950")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, xujxyang/fairness-nc-main 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.

process_batch Ran
ran; our contract draft was wrong, not the code
Xujxyang/Fairness-NC-main/adept/debias.py
pointer only (licence: NONE) · get_code("d3c361e2a27ed6ed")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Xujxyang/Fairness-NC-main 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.

process_batch Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Xujxyang/Fairness-NC-main/adept/lib/collection.py
pointer only (licence: NONE) · get_code("8ef531ec328be202")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Xujxyang/Fairness-NC-main 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.

split_data Ran
ran; our contract draft was wrong, not the code
Xujxyang/Fairness-NC-main/adept/debias_original.py
pointer only (licence: NONE) · get_code("d0dd7e697cd0cfba")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Xujxyang/Fairness-NC-main 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.

truncate_and_pad Ran
ran; our contract draft was wrong, not the code
Xujxyang/Fairness-NC-main/adept/lib/collection.py
pointer only (licence: NONE) · get_code("cc361c2f7b68b11d")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Xujxyang/Fairness-NC-main 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.

variable Ran
ran; our contract draft was wrong, not the code
Xujxyang/Fairness-NC-main/ase/model.py
pointer only (licence: NONE) · get_code("32967e6aa4441efb")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Xujxyang/Fairness-NC-main 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.

prepare_tokenizer Not yet run
unverified (harvested, no recorded run)
Xujxyang/Fairness-NC-main/adept/collect_sentences.py
pointer only (licence: NONE) · get_code("42fe7317678380ab")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Xujxyang/Fairness-NC-main 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.

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 12 distinct functions with an identity, 9 are also present in other papers whose code we hold (9 with at least one paper dated earlier, 0 the same month, 0 later) and 3 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
variable 4 other papers
4 dated earlier
1711.04436 (2017-11)
SQLNet: Generating Structured Queries From Natural Language Without Reinforcement Learning
2004.14366 (2020-04)
Elastic weight consolidation for better bias inoculation
2103.12827 (2021-03)
Fisher Task Distance and Its Application in Neural Architecture Search
and 1 more not named here
create_dataset 2 other papers
2 dated earlier
2101.09523 (2021-01)
Debiasing Pre-trained Contextualised Embeddings
2211.05414 (2022-11)
ADEPT: A DEbiasing PrompT Framework
split_data 2 other papers
2 dated earlier
2101.09523 (2021-01)
Debiasing Pre-trained Contextualised Embeddings
2211.05414 (2022-11)
ADEPT: A DEbiasing PrompT Framework
calculate_group_to_group_relative_distance_asymmetric 1 other paper
1 dated earlier
2211.05414 (2022-11)
ADEPT: A DEbiasing PrompT Framework
calculate_group_to_group_relative_distance_asymmetric_test 1 other paper
1 dated earlier
2211.05414 (2022-11)
ADEPT: A DEbiasing PrompT Framework
calculate_group_to_one_relative_distance_asymmetric_test 1 other paper
1 dated earlier
2211.05414 (2022-11)
ADEPT: A DEbiasing PrompT Framework
load_and_cache_examples 1 other paper
1 dated earlier
2211.05414 (2022-11)
ADEPT: A DEbiasing PrompT Framework
process_batch 1 other paper
1 dated earlier
2405.17767 (2024-05)
Linguistic Collapse: Neural Collapse in (Large) Language Models
truncate_and_pad 1 other paper
1 dated earlier
2405.17767 (2024-05)
Linguistic Collapse: Neural Collapse in (Large) Language Models

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

To mitigate societal biases implicitly encoded in recent successful pretrained language models, a diverse array of approaches have been proposed to encourage model fairness, focusing on prompting, data augmentation, regularized finetuning, and more. Despite the development, it is nontrivial to reach a principled understanding of fairness and an effective algorithm that can consistently debias language models. In this work, by rigorous evaluations of Neural Collapse -a learning phenomenon happen in last-layer representations and classifiers in deep networks -on fairness-related words, we find that debiased language models exhibit collapsed alignment between token representations and word embeddings. More importantly, this observation inspires us to design a principled fine-tuning method that can effectively improve fairness in a wide range of debiasing methods, while still preserving the performance of language models on standard natural language understanding tasks. We attach our code at https://github.com/Xujxyang/Fairness-NC-main.

For agents

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

get_harvested_code_for_paper("2410.04472")
get_code_for_paper("2410.04472")
have("2410.04472")

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

curl https://syntology.ai/api/ran/2410.04472.json

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

[![Syntology run record](https://syntology.ai/api/ran/2410.04472.svg)](https://syntology.ai/paper/2410.04472)

Connect an agent — have() is free.