Mitigating Biases of Large Language Models in Stance Detection with Calibration
Li, Ang, Zhao, Jingqian, Liang, Bin, Gui, Lin, Wang, Hui, Zeng, Xi, Liang, Xingwei, Wong, Kam-Fai, Xu, Ruifeng
–arXiv.org Artificial Intelligence
Large language models (LLMs) have achieved remarkable progress in many natural language processing tasks. However, our experiment reveals that, in stance detection tasks, LLMs may generate biased stances due to sentiment-stance spurious correlations and preference towards certain individuals and topics, thus harming their performance. Therefore, in this paper, we propose to Mitigate Biases of LLMs in stance detection with Calibration (MB-Cal). To be specific, a novel calibration network is devised to calibrate potential bias in the stance prediction of LLMs. Further, to address the challenge of effectively learning bias representations and the difficulty in the generalizability of debiasing, we construct counterfactual augmented data. This approach enhances the calibration network, facilitating the debiasing and out-of-domain generalization. Experimental results on in-target and zero-shot stance detection tasks show that the proposed MB-Cal can effectively mitigate biases of LLMs, achieving state-of-the-art results.
arXiv.org Artificial Intelligence
Jun-16-2024
- Country:
- Asia > China (0.68)
- Europe (1.00)
- North America > United States
- New Mexico (0.14)
- Genre:
- Research Report > New Finding (0.46)
- Industry:
- Technology: