Although Large Language Models (LLMs) perform exceptionally well in general domains, the problem of hallucinations poses significant risks in specialized fields such as healthcare and law, where high interpretability is essential. Existing fine-tuning methods depend heavily on large-scale professional datasets, which are often hard to obtain due to the privacy regulations. Moreover, existing self-evolution methods are primarily designed for general domains, which may struggle to adapt to knowledge-intensive domains due to the lack of knowledge constraints. In this paper, we propose an ontology rule guided method Evontree to enable self-evolution of LLMs in low-resource specialized domains. Specifically, Evontree first extracts domain ontology knowledge from raw models, then detects knowledge inconsistencies using two core ontology rules, and finally reinforces gap knowledge into model via self-distilled fine-tuning. Extensive evaluations on medical QA benchmarks using Llama3-8B-Instruct and Med42-V2 demonstrate the effectiveness of Evontree, which outperforms both the base models and strong baselines, achieving up to a 3.7\% improvement in accuracy. Detailed ablation studies further validate the robustness of our approach.
TextMineX: Data, Evaluation Framework and Ontology-guided LLM Pipeline for Humanitarian Mine Action
NEC Laboratories Europe(NEC欧洲实验室);University of Stuttgart(斯图加特大学);VAGO Solutions(VAGO解决方案);Zurich University of Applied Sciences(苏黎世应用科学大学);CAIR, Ss. Cyril and Methodius University of Skopje, North Macedonia(CAIR,斯科普耶塞尔维亚·梅托迪乌斯大学,北马其顿)
Ontology-Guided Reverse Thinking Makes Large Language Models Stronger on Knowledge Graph Question Answering
Harbin Institute of Technology(哈尔滨工业大学);Beijing University of Posts and Telecommunications(北京邮电大学);Joint Laboratory of HIT and iFLYTEK(哈工大与科大讯飞联合实验室);University of Science and Technology of China(中国科学技术大学);Pengcheng Laboratory(鹏城实验室)
Feasibility of Structuring Stress Documentation Using an Ontology-Guided Large Language Model
College of Nursing, Seoul National University(首尔国立大学护理学院);The Research Institute of Nursing Science, Seoul National University(首尔国立大学护理科学研究所);Center of Human-Caring Nurse Leaders for the Future by the Brain Korea (BK 21) four project, College of Nursing, Seoul National University(首尔国立大学护理学院,韩国脑计划(BK21)第四阶段未来人文关怀护理领军人才中心);College of Medicine, Seoul National University(首尔国立大学医学院)
Concept-Guided Noisy Negative Suppression for Zero-Shot Classification and Grounding of Chest X-Ray Findings
The Center for Smart Health, School of Nursing, the Hong Kong Polytechnic University, Hong Kong, China(香港理工大学智能健康中心,护理学院,中国香港);Research Institute for Smart Ageing, the Hong Kong Polytechnic University, Hong Kong, China(香港理工大学智能老龄化研究 institute,中国香港);School of Biomedical Engineering, Tsinghua Medicine, Tsinghua University, Beijing, China(清华大学生物医学工程学院,清华大学,北京,中国);Queen Mary Hospital, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong, China(香港大学李嘉诚医学院Queen Mary医院,中国香港)
AOR: Anatomical Ontology-Guided Reasoning for Medical Large Multimodal Model in Chest X-Ray Interpretation
Fudan University(复旦大学);Xidian University(西安电子科技大学);KAIST(韩国科学技术院);Children’s Hospital of Fudan University(复旦大学附属儿童医院);Shanghai AI Laboratory(上海人工智能实验室);Hong Kong Polytechnic University(香港理工大学)
Manual-Guided Dialogue for Flexible Conversational Agents
Tsinghua University(清华大学);Tencent Inc.(腾讯公司);Beijing National Research Center for Information Science and Technology(北京信息科学与技术国家研究中心);WeChat AI, Tencent Inc.(腾讯微信人工智能部)
Explaining, Verifying, and Aligning Semantic Hierarchies in Vision-Language Model Embeddings
University of Lübeck(吕贝克大学);Technical University of Munich(慕尼黑工业大学);AUMOVIO SE;Osnabrück University(奥斯纳布吕克大学);Leibniz Institute for Agricultural Engineering and Bioeconomy(莱布尼茨农业工程与生物经济研究所);University of Hamburg(汉堡大学)
Towards Scalable Multi-domain Conversational Agents: The Schema-Guided Dialogue Dataset
Google Research(谷歌研究院)
Google Assistant、Alexa和Siri等虚拟助理为跨多个域的大量服务和API提供对话界面。这样的系统需要支持不断增加的服务数量,这些服务可能具有重叠的功能。此外,这些服务中的一些服务几乎没有可用的训练数据。现有的面向任务的对话的公共数据集不足以捕捉这些挑战,因为它们覆盖几个领域,并假设每个域的一个静态本体。在这项工作中,我们介绍了模式引导对话(SGD)数据集,包含超过16k的多域会话跨越16个域。我们的数据集在规模上超过了现有的面向任务的对话语料库,同时也突出了与构建大规模虚拟助手相关的挑战。它为许多任务提供了一个具有挑战性的测试平台,包括语言理解、槽填充、对话状态跟踪和响应生成。沿着同样的思路,我们提出了一个模式引导的范式为任务导向的对话,其中预测的意图和插槽,作为输入,使用他们的自然语言描述的动态设置。这允许单个对话系统容易地支持大量服务,并且便于新服务的简单集成,而不需要额外的训练数据。基于所提出的范例,我们发布了一个对话状态跟踪模型,能够zero-shot推广到新的API,同时在常规设置中保持竞争力。