From Pretraining Data to Language Models to Downstream Tasks: Tracking the Trails of Political Biases Leading to Unfair NLP Models
By Shangbin Feng, Chan Young Park, Yuhan Liu, Yulia Tsvetkov. Annual Meeting of the Association for Computational Linguistics, 2023
396 citations recorded by the source corpus.
Language models (LMs) are pretrained on diverse data sources—news, discussion forums, books, online encyclopedias. A significant portion of this data includes facts and opinions which, on one hand, celebrate democracy and diversity of ideas, and on the other hand are inherently socially biased. Our work develops new methods to (1) measure media biases in LMs trained on such corpora, along social and economic axes, and (2) measure the fairness of downstream NLP models trained on top of politically biased LMs. We focus on hate speech and misinformation detection, aiming to empirically quantify the effects of political (social, economic) biases in pretraining data on the fairness of high-stakes social-oriented tasks. Our findings reveal that pretrained LMs do have political leanings which reinforce the polarization present in pretraining corpora, propagating social biases into hate speech predictions and media biases into misinformation detectors. We discuss the implications of our findings for NLP research and propose future directions to mitigate unfairness.
Subfields: Artificial Intelligence
View Paper: https://www.semanticscholar.org/paper/5471114e37448bea2457b74894b1ecb92bbcfdf6
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Karthik Narasimhan
- Colin Raffel
- Mohit Bansal
- Shunyu Yao
- Thomas Wolf
- Theodore R. Sumers
- Thomas L. Griffiths
- Jianfeng Gao
Citations in the corpus
- Cognitive Architectures for Language Agents (2023, background): “More capable agents (e.g., Voyager, Generative Agents) have larger action spaces – which in turn means they face a more complex decision-making problem.”
- The FineWeb Datasets: Decanting the Web for the Finest Text Data at Scale (2024, background): “Language models are known to reflect the biases present in their pretraining datasts [67–72].”