Jiaxu Zuo
Research Subjects
Jiaxu Zuo has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Jing Zhang
Universidade de Macau
3 shared publications - Mu You
University of Macau
3 shared publications - Kaixin Lan
University of Macau
2 shared publications - Lidia Sam Chao
University of Macau
2 shared publications - Tao Fang
Guangdong Institute of Intelligence Science and Technology, Zhuhai, Guangdong, China
2 shared publications - Ana Cristina Ferreira de Almeida Rodrigues Alves
Macau Scientific and Cultural Centre, Portugal
1 shared publication
Latest Publications
- 29 July 2026
- Published by Elsevier in Expert Systems with Applications
- Vol. 333
Masked cloze scoring (MCS): Zero-shot automated essay scoring via masked predictions from large language models
- 29 July 2026
- Published by Elsevier in Expert Systems with Applications
- Vol. 333
Masked cloze scoring (MCS): Zero-shot automated essay scoring via masked predictions from large language models
- Open Access
- 14 February 2026
- Published by Elsevier in SSRN Electronic Journal
Masked Cloze Scoring (MCS): Zero-Shot Automated Essay Scoring via Masked Predictions from Large Language Models
Traditional Automated Essay Scoring (AES) methods typically require large amounts of labeled data and extensive model training. Recent advances in Large Language Models (LLMs) have enabled zero-shot AES methods that bypass these requirements. In this paper, we propose Masked Cloze Scoring (MCS), a novel zero-shot method that leverages LLMs’ masked language modeling capacity as its core mechanism for evaluating how well an essay aligns with model predictions. Unlike prior methods that prompt LLMs to assign scores directly, MCS integrates two complementary signals: in-context word prediction for micro-level lexical appropriateness, and relative essay length as a proxy for macro-level writing quality. Experimental results demonstrate that MCS consistently outperforms existing training-free zero-shot baselines on two widely used benchmark datasets, ASAP and TOEFL11. Compared with LCES, the state-of-the-art zero-shot method that requires additional training, MCS underperforms on TOEFL11 but achieves comparable performance on ASAP. Moreover, MCS attains performance comparable to that of three strong supervised cross-prompt baselines on ASAP. Beyond holistic scoring, MCS produces word-level diagnostic feedback, offering added pedagogical benefits. Overall, MCS advances the state of training-free zero-shot AES, delivering competitive performance without requiring labeled data or additional training.
- 14 February 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessMasked Cloze Scoring (MCS): Zero-Shot Automated Essay Scoring via Masked Predictions from Large Language Models
Traditional Automated Essay Scoring (AES) methods typically require large amounts of labeled data and extensive model training. Recent advances in Large Language Models (LLMs) have enabled zero-shot AES methods that bypass these requirements. In this paper, we propose Masked Cloze Scoring (MCS), a novel zero-shot method that leverages LLMs’ masked language modeling capacity as its core mechanism for evaluating how well an essay aligns with model predictions. Unlike prior methods that prompt LLMs to assign scores directly, MCS integrates two complementary signals: in-context word prediction for micro-level lexical appropriateness, and relative essay length as a proxy for macro-level writing quality. Experimental results demonstrate that MCS consistently outperforms existing training-free zero-shot baselines on two widely used benchmark datasets, ASAP and TOEFL11. Compared with LCES, the state-of-the-art zero-shot method that requires additional training, MCS underperforms on TOEFL11 but achieves comparable performance on ASAP. Moreover, MCS attains performance comparable to that of three strong supervised cross-prompt baselines on ASAP. Beyond holistic scoring, MCS produces word-level diagnostic feedback, offering added pedagogical benefits. Overall, MCS advances the state of training-free zero-shot AES, delivering competitive performance without requiring labeled data or additional training.
- Open Access
- 1 January 2026
- Published by FapUNIFESP (SciELO) in Texto Livre
- Vol. 19
Exploring learner prompting and AI feedback quality in L2 Portuguese writing
For L2 Portuguese learners in Macau, artificial intelligence (AI) has become a primary source of immediate writing feedback for Chinese university students. This study investigates authentic learner prompting practices and the accuracy of resulting AI-generated revisions in L2 Portuguese writing. Forty-six second-year Portuguese Studies majors, divided into a control condition and a justification condition that required them to rationalize every modification, revised a first-year learner composition using their preferred AI tools. Results reveal that students predominantly prompted in Chinese, employed highly anthropomorphic and polite language, relied on generic or grammar-focused requests, and frequently exhibited misalignment between intended goals and actual prompts. The requirement to justify revisions increased interaction turns and shifted attention toward smaller textual units and explanation-seeking, though it also produced higher error rates in the resulting revisions. Experiments with researcher-designed prompts demonstrated the potential of simple prompt engineering strategies, with the Rephrase-and-Respond technique eliciting the most accurate revisions. The findings underscore the immaturity of untrained prompting, the context-sensitive nature of prompt engineering, and the practical pedagogical value of current LLMs even under naïve use. To foster critical and reflective human-AI collaboration in L2 writing, pedagogical interventions such as explicit instruction in prompting strategies and the use of tasks that require justification or explanation-seeking are recommended.
- 1 January 2026
- Published by FapUNIFESP (SciELO) in Texto Livre
- Vol. 19
Open AccessExploring learner prompting and AI feedback quality in L2 Portuguese writing
For L2 Portuguese learners in Macau, artificial intelligence (AI) has become a primary source of immediate writing feedback for Chinese university students. This study investigates authentic learner prompting practices and the accuracy of resulting AI-generated revisions in L2 Portuguese writing. Forty-six second-year Portuguese Studies majors, divided into a control condition and a justification condition that required them to rationalize every modification, revised a first-year learner composition using their preferred AI tools. Results reveal that students predominantly prompted in Chinese, employed highly anthropomorphic and polite language, relied on generic or grammar-focused requests, and frequently exhibited misalignment between intended goals and actual prompts. The requirement to justify revisions increased interaction turns and shifted attention toward smaller textual units and explanation-seeking, though it also produced higher error rates in the resulting revisions. Experiments with researcher-designed prompts demonstrated the potential of simple prompt engineering strategies, with the Rephrase-and-Respond technique eliciting the most accurate revisions. The findings underscore the immaturity of untrained prompting, the context-sensitive nature of prompt engineering, and the practical pedagogical value of current LLMs even under naïve use. To foster critical and reflective human-AI collaboration in L2 writing, pedagogical interventions such as explicit instruction in prompting strategies and the use of tasks that require justification or explanation-seeking are recommended.