English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Recent literature has suggested the potential of using large language models(LLMs) to make predictions for tabular tasks. However, LLMs have been shown toexhibit harmful social biases that reflect the stereotypes and inequalitiespresent in the society. To this end, as well as the widespread use of tabulardata in many high-stake applications, it is imperative to explore the followingquestions: what sources of information do LLMs draw upon when makingpredictions for tabular tasks; whether and to what extent are LLM predictionsfor tabular tasks influenced by social biases and stereotypes; and what are theconsequential implications for fairness? Through a series of experiments, wedelve into these questions and show that LLMs tend to inherit social biasesfrom their training data which significantly impact their fairness in tabularprediction tasks. Furthermore, our investigations show that in the context ofbias mitigation, though in-context learning and fine-tuning have a moderateeffect, the fairness metric gap between different subgroups is still largerthan that in traditional machine learning models, such as Random Forest andshallow Neural Networks. This observation emphasizes that the social biases areinherent within the LLMs themselves and inherited from their pre-trainingcorpus, not only from the downstream task datasets. Besides, we demonstratethat label-flipping of in-context examples can significantly reduce biases,further highlighting the presence of inherent bias within LLMs.

Investigating the Fairness of Large Language Models for Predictions on Tabular Data