We test whether large audio language models use prosodic and semantic cues to resolve relative-clause attachment ambiguities in English, Chinese, and Arabic. Models show limited sensitivity to prosody but greater reliance on semantic and world-knowledge cues.
@inproceedings{hu2026relative,title={Where Does the Relative Clause Attach? Probing Prosodic and Semantic Sensitivity in Large Audio Language Models},author={Hu, Jingying and Issa, Elsayed},booktitle={Conference on Language Modeling},year={2026},}
C&E
LLM-derived metrics in second language writing assessment: an explainable AI approach
We use LLM-derived surprisal, perplexity, and embedding similarity as interpretable measures of linguistic predictability and semantic coherence in Chinese L2 writing. These metrics track proficiency and complement classical linguistic features, improving proficiency assessment when combined.
@article{hu2026llm,title={LLM-derived metrics in second language writing assessment: an explainable AI approach},author={Hu, Jingying and Cong, Yan},journal={Computers \& Education},year={2026},pages={105721},publisher={Elsevier},doi={10.1016/j.compedu.2026.105721},}
NLPJ
How robust are linguistic markers of aging? The case of aging-related social media text
This study examines whether linguistic biomarkers validated in clinical settings can robustly detect probable cognitive impairment (PCI) in unstructured social media posts. By combining classic linguistic features with LLM-derived surprisal and similarity measures, we construct machine learning models and evaluate zero-shot LLM prompting, demonstrating robust cross-domain performance and the potential for early detection in everyday social media language.
@article{cong2026aging,title={How robust are linguistic markers of aging? The case of aging-related social media text},author={Cong, Yan and Hu, Jingying and Reese, Timothy and Liu, Hui},journal={Natural Language Processing Journal},volume={14},year={2026},pages={100203},publisher={Elsevier},doi={10.1016/j.nlp.2026.100203},}
2025
CMCL
Modeling Chinese L2 Writing Development: The LLM-Surprisal Perspective
Jingying Hu and Yan Cong
In Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics, 2025
This study explores how LLM can inform language assessment for second language (L2) learners. Specifically, we examine LLM-derived surprisal as a scalable and interpretable metric for modeling Chinese L2 writing development, demonstrating its potential as a cross-linguistically robust measure for automatic writing assessment.
@inproceedings{hu2025modeling,title={Modeling Chinese L2 Writing Development: The LLM-Surprisal Perspective},author={Hu, Jingying and Cong, Yan},booktitle={Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics},year={2025},pages={172--183},address={Albuquerque, New Mexico, USA},publisher={Association for Computational Linguistics},doi={10.18653/v1/2025.cmcl-1.22},}
2023
Front Psychol
The influence of metacognition monitoring on L2 Chinese audiovisual reading comprehension
Yamin Wang, Jingying Hu, Zhuoma An, Chaoran Li, and Yang Zhao
We investigate how different dimensions of metacognitive monitoring predict L2 Chinese audiovisual comprehension. Absolute calibration accuracy predicts comprehension whereas relative calibration accuracy does not, with the predictive effects varying by video difficulty and learners’ L2 Chinese proficiency
@article{wang2023metacognition,title={The influence of metacognition monitoring on L2 Chinese audiovisual reading comprehension},author={Wang, Yamin and Hu, Jingying and An, Zhuoma and Li, Chaoran and Zhao, Yang},journal={Frontiers in Psychology},volume={14},year={2023},pages={1133003},doi={10.3389/fpsyg.2023.1133003},}
