Papers by Yu-Jie Xiong
From Answers to Arguments: Toward Trustworthy Clinical Diagnostic Reasoning with Toulmin-Guided Curriculum Goal-Conditioned Learning (2026.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are obstructed by their opaque and often unreliable reasoning. |
| Approach: | They propose a framework for trustworthy clinical argumentation by adapting the Toulmin model to the diagnostic process. |
| Outcome: | The proposed method achieves diagnostic accuracy comparable to resource-intensive RL methods while offering a more stable and efficient training pipeline. |
Parameter-Efficient Fine-Tuning of Large Language Models via Deconvolution in Subspace (2025.coling-main)
Copied to clipboard
| Challenge: | Existing methods for parameter-efficient fine-tuning have been proposed to reduce time and resource costs. |
| Approach: | They propose a parameter-efficient fine-tuning method that combines the knowledge completion capability of deconvolution with the subspace learning ability, reducing the number of parameters required for fine-uning by 8 times. |
| Outcome: | The proposed method reduces the number of parameters required for fine-tuning by 8 times and achieves comparable or superior performance compared to existing models. |
Sugar-Coated Poison: Benign Generation Unlocks Jailbreaking (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods to jailbreak large language models rely on black-box manipulation of prompt templates, resulting in high costs and poor generalizability. |
| Approach: | They propose a sugar-coated poison attack paradigm that uses a "semantic reversal" strategy to induce the model into a safety response mode. |
| Outcome: | The proposed attack paradigm outperforms baselines in the study. |