Challenge: Large language models acquire general knowledge from pretraining but pretraining data contain undesirable social biases which can be perpetuated or even amplified by LLMs.
Approach: They propose an efficient yet effective annotation pipeline to investigate social biases in pretraining data.
Outcome: The proposed pipeline investigates social biases in the pretraining corpus using protected attribute detection and regard classification.

Similar Papers

A Predictive Factor Analysis of Social Biases and Task-Performance in Pretrained Masked Language Models (2023.emnlp-main)

Copied to clipboard

Challenge: Various types of social biases have been reported with pretrained Masked Language Models (MLMs) in prior work.
Approach: They conduct a comprehensive study on 39 pretrained MLMs to examine their model factors and their social biases.
Outcome: The proposed model factors influence social biases learned by an MLM and their downstream task performance.
From Pretraining Data to Language Models to Downstream Tasks: Tracking the Trails of Political Biases Leading to Unfair NLP Models (2023.acl-long)

Copied to clipboard

Challenge: Hundreds of studies have highlighted ethical issues in NLP models .
Approach: They propose to measure media biases in LMs trained on diverse data sources . they focus on hate speech and misinformation detection .
Outcome: The proposed methods quantify the fairness of downstream NLP models trained on politically biased LMs.
Social Bias in Multilingual Language Models: A Survey (2025.emnlp-main)

Copied to clipboard

Challenge: Pretrained multilingual models exhibit the same social bias as models processing English texts.
Approach: They examine the literature on bias evaluation and mitigation approaches in multilingual and non-English contexts and identify gaps in the field.
Outcome: The proposed models perform well on multilingual language understanding benchmarks and are consistent with the current literature.
Debiasing Large Language Models with Structured Knowledge (2024.findings-acl)

Copied to clipboard

Challenge: Existing methods to reduce biases in pre-training models are hampered by their performance.
Approach: They propose a method that utilizes structured knowledge to mitigate bias in LLMs . their method obviates the need for training from scratch, thus offering enhanced scalability .
Outcome: The proposed method outperforms state-of-the-art (SOTA) baselines in the debiasing ability.
Understanding the Effect of Model Compression on Social Bias in Large Language Models (2023.emnlp-main)

Copied to clipboard

Challenge: Large Language Models (LLMs) trained with self-supervision on vast corpora of web text fit to the social biases of that text, leading to representational harm.
Approach: They propose to use quantization and knowledge distillation to reduce the computational burden of LLMs to mitigate the effects of inappropriate social biases learned during pretraining.
Outcome: The proposed methods reduce the computational burden of large language models by reducing their size and complexity.
Deciphering the Impact of Pretraining Data on Large Language Models through Machine Unlearning (2024.findings-acl)

Copied to clipboard

Challenge: Existing studies have suggested that the composition of the pretraining corpus exerts a significant impact upon the performance of LLMs.
Approach: They analyze the impact of 48 datasets from 5 major categories of pretraining data of Large Language Models and measure their impacts on LLMs using benchmarks about nine major categories.
Outcome: The proposed analysis provides insights into the organization of data to support more efficient pretraining of Large Language Models.
CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Models (2020.emnlp-main)

Copied to clipboard

Challenge: Pretrained language models use cultural biases implicitly, causing harm . identifying and quantifying learnt biase enables us to measure progress .
Approach: They propose a benchmark to measure social bias in pretrained language models . they use 1508 examples that cover stereotypes dealing with nine types of bias .
Outcome: The proposed benchmark focuses on stereotypes about historically disadvantaged groups and contrasts them with advantaged groups.
BiasDPO: Mitigating Bias in Language Models through Direct Preference Optimization (2024.acl-srw)

Copied to clipboard

Challenge: Large Language Models (LLMs) have been shown to be effective in complex language tasks, but their potential to perpetuate biases poses significant concerns.
Approach: They propose a new framework employing Direct Preference Optimization to mitigate biases in LLMs.
Outcome: The proposed model outperforms the baseline model on almost all bias benchmarks and achieves better performance than open-source models.
A Study of Implicit Bias in Pretrained Language Models against People with Disabilities (2022.coling-1)

Copied to clipboard

Challenge: Pretrained language models exhibit sociodemographic biases, such as against gender and race, raising concerns of downstream biase in language technologies.
Approach: They propose to use word embedding-based and transformer-based PLMs to test for the presence of biases against people with disabilities (PWDs)
Outcome: The proposed models favor ableist language, despite their sociodemographic biases against race and gender.
Biases in Large Language Model-Elicited Text: A Case Study in Natural Language Inference (2025.coling-main)

Copied to clipboard

Challenge: Creating NLP datasets with Large Language Models (LLMs) is an attractive alternative to relying on crowd-source workers.
Approach: They recreate a portion of the Stanford Natural Language Inference corpus using GPT-4, Llama-2 70b for Chat, and Mistral 7b Instruct.
Outcome: The proposed model can be used to generate NLP datasets with stereotypical biases and annotation artifacts.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations