Papers by Zhuangdi Zhu
Dialogue is Better Than Monologue: Instructing Meidcal LLMs via Strategic Conversations (2026.findings-eacl)
Copied to clipboard
Zijie Liu, Xinyu Zhao, Jie Peng, Jinhao Duan, Zhuangdi Zhu, Qingyu Chen, Kaidi Xu, Xia Hu, Tianlong Chen
| Challenge: | Existing tuning methods for medical AI models are monologue-based . existing benchmarks are based on licensing exams or research articles . |
| Approach: | They propose a benchmark to expose limitations of monologue-based tuning for medical AI models . they use a large dialogue dataset to capture stepwise diagnostic reasoning . |
| Outcome: | The proposed model outperforms monologue-tuned models on a medical question answering task and improves accuracy on standard medical QA benchmarks. |
Web Intellectual Property at Risk: Preventing Unauthorized Real-Time Retrieval by Large Language Models (2025.emnlp-main)
Copied to clipboard
| Challenge: | a new framework protects web content from unauthorized LLM real-time extraction and redistribution . multiple AI companies have been accused of scraping digital IP for proprietary benefit . |
| Approach: | They propose a defense framework that empowers web content creators to safeguard their web-based IP from unauthorized LLM real-time extraction and redistribution by leveraging the semantic understanding capability of LLMs themselves. |
| Outcome: | The proposed defense outperforms traditional defenses on LLMs and improves on black-box optimization problems. |