Li Zhu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Li Zhu has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
- Hyesun You
Science Education, Department of Teaching and Learning, College of Education, The University of Iowa, Iowa City, USA
3 shared publications - Minju Hong
Department of Psychology, Chung-Ang University, Seoul, Korea
2 shared publications Ariel M Aloe
University of Iowa (United States, Iowa City) - UI
1 shared publication- Fang Zhenhan
Department of Statistics and Actuarial Science, College of Liberal Arts and Sciences, The University of Iowa, Iowa City, USA
1 shared publication Kishlay Jha
University of Iowa (United States, Iowa City) - UI
1 shared publicationZhenhan Fang
1 shared publication
Latest Publications
- Open Access
- 26 October 2026
- Published by The University of Iowa
Fine-Tuning Large Language Models for Text Summarization
Large language models (LLMs) have shown strong potential for abstractive summarization, yet most prior work in evidence synthesis limits summarization to titles and abstracts and primarily focuses on domains with well-structured reporting formats. These limitations restrict the applicability of automated PICO summarization in social science research, where key information is often embedded in unstructured full-text sections. This study presents a comprehensive end-to-end workflow for fine-tuning LLMs for PICO-oriented summarization of long-form scientific texts in the mental health domain. The proposed pipeline integrates document-level annotation, label-specific instance construction, section-aware data augmentation, long-context model architectures, optimization strategies, and multi-metric evaluation within a unified framework. To systematically examine the effects of model and training choices, we employ a 2×3×3 factorial design varying model architecture, batch size, and learning rate. Performance is evaluated using three complementary metrics (ROUGE, perplexity, MAUVE), supplemented by qualitative comparisons between model-generated and human-written summaries. Results show that LongT5 consistently achieves higher ROUGE and lower perplexity, while BioBART attains higher MAUVE. Across configurations, intermediate learning rates and moderate batch sizes yield the most balanced performance. Further analysis reveals that higher learning rates disproportionately benefit structured and quantitative categories, while gains are more limited for descriptive categories. These findings highlight systematic trade-offs between accuracy, confidence, and diversity in generated summaries. Overall, the proposed pipeline provides practical, evidence-based guidance for applying generative models in long-document evidence synthesis and offers a foundation for future research on finetuning LLMs.
- 26 October 2026
- Published by The University of Iowa
Open AccessFine-Tuning Large Language Models for Text Summarization
Large language models (LLMs) have shown strong potential for abstractive summarization, yet most prior work in evidence synthesis limits summarization to titles and abstracts and primarily focuses on domains with well-structured reporting formats. These limitations restrict the applicability of automated PICO summarization in social science research, where key information is often embedded in unstructured full-text sections. This study presents a comprehensive end-to-end workflow for fine-tuning LLMs for PICO-oriented summarization of long-form scientific texts in the mental health domain. The proposed pipeline integrates document-level annotation, label-specific instance construction, section-aware data augmentation, long-context model architectures, optimization strategies, and multi-metric evaluation within a unified framework. To systematically examine the effects of model and training choices, we employ a 2×3×3 factorial design varying model architecture, batch size, and learning rate. Performance is evaluated using three complementary metrics (ROUGE, perplexity, MAUVE), supplemented by qualitative comparisons between model-generated and human-written summaries. Results show that LongT5 consistently achieves higher ROUGE and lower perplexity, while BioBART attains higher MAUVE. Across configurations, intermediate learning rates and moderate batch sizes yield the most balanced performance. Further analysis reveals that higher learning rates disproportionately benefit structured and quantitative categories, while gains are more limited for descriptive categories. These findings highlight systematic trade-offs between accuracy, confidence, and diversity in generated summaries. Overall, the proposed pipeline provides practical, evidence-based guidance for applying generative models in long-document evidence synthesis and offers a foundation for future research on finetuning LLMs.
- 12 October 2026
- Published by Taylor & Francis in Journal of College Science Teaching
- p. 1-11
Developing a Custom GPT for NGSS-PLUS-5E Lesson Planning in Elementary Science Education
A lesson plan is a teacher’s comprehensive roadmap that aligns specific learning objectives with disciplinary content, instructional strategies, and assessment methods to support organized teaching...
- 12 October 2026
- Published by Taylor & Francis in Journal of College Science Teaching
- p. 1-11
Developing a Custom GPT for NGSS-PLUS-5E Lesson Planning in Elementary Science Education
A lesson plan is a teacher’s comprehensive roadmap that aligns specific learning objectives with disciplinary content, instructional strategies, and assessment methods to support organized teaching...