TuiML: Machine Learning for AI Agents
为解决AI代理使用传统机器学习库时遇到的问题，本文提出TuiML，一个专为AI代理设计的自包含机器学习库，通过可机读元数据和参数模式提高搜索、验证及实验再现性。
0 citationsRead paper
Selected work
Representative Papers
Evaluating and Preventing Security Smells in AI-Generated Ansible Code
本文研究了AI生成的Ansible代码是否满足安全要求，并提出了一种将Ansible最佳实践和CIS基准集成到提示中的方法，以预防安全问题。
0 citationsRead paper
Data Findability, Governance, and Community Engagement for Māori Research Data Sovereignty
This study addresses how research institutions can operationalize Māori data sovereignty principles throughout the full data lifecycle to uphold Māori rights and interests in their data. It introduces, for the first time, the concept of “Māori Research Data Sovereignty” (MRDSov) and develops an actionable framework that integrates Indigenous data sovereignty theory with research data management practice. Centered on three interdependent pillars—data discoverability, governance mechanisms, and community engagement—the framework eschews reliance on specific technical tools in favor of institutional pathways. It enables universities and research organizations to fulfill their obligations under the Treaty of Waitangi by embedding Māori authority, accountability, and decision-making power into research data processes.
0 citationsRead paper
Rashomon Alignment
Existing approaches to measuring functional similarity between models rely on the true data distribution, making it difficult to characterize alignment of decision boundaries across the entire input space. This work proposes Rashomon Alignment (RA), a novel framework that, for the first time, evaluates functional similarity between models from a geometric perspective over the full input space without dependence on any specific data distribution. By uniformly sampling the input space and employing geometric similarity metrics, RA enables a global analysis of decision boundary alignment. Experiments across more than 90 datasets demonstrate that geometric alignment provides a complementary perspective to distribution-based alignment, and that RA effectively supports model selection, ensemble construction, and enhanced interpretability.
0 citationsRead paper
Does Role Specialization Matter for Explanation Faithfulness in Mixture-of-Experts?
This work addresses the limited faithfulness of explanations in existing Mixture-of-Experts (MoE) models, which stems primarily from high representational overlap among experts that undermines role specialization. To mitigate this issue, the authors propose a representation-level decorrelation regularization mechanism that explicitly reduces the similarity between experts’ latent representations, thereby enhancing their functional distinctiveness. The method consistently improves key faithfulness metrics—including comprehensiveness, sufficiency, and AOPC—across multiple multimodal benchmarks without compromising task performance. Furthermore, the approach demonstrates broad applicability and effectiveness within standard sparse MoE architectures, validating its potential as a general-purpose enhancement for interpretable expert-based models.
0 citationsRead paper
Recent publications
Latest Papers
TuiML: Machine Learning for AI Agents
为解决AI代理使用传统机器学习库时遇到的问题，本文提出TuiML，一个专为AI代理设计的自包含机器学习库，通过可机读元数据和参数模式提高搜索、验证及实验再现性。
0 citationsRead paper
Evaluating and Preventing Security Smells in AI-Generated Ansible Code
本文研究了AI生成的Ansible代码是否满足安全要求，并提出了一种将Ansible最佳实践和CIS基准集成到提示中的方法，以预防安全问题。
0 citationsRead paper
Data Findability, Governance, and Community Engagement for Māori Research Data Sovereignty
This study addresses how research institutions can operationalize Māori data sovereignty principles throughout the full data lifecycle to uphold Māori rights and interests in their data. It introduces, for the first time, the concept of “Māori Research Data Sovereignty” (MRDSov) and develops an actionable framework that integrates Indigenous data sovereignty theory with research data management practice. Centered on three interdependent pillars—data discoverability, governance mechanisms, and community engagement—the framework eschews reliance on specific technical tools in favor of institutional pathways. It enables universities and research organizations to fulfill their obligations under the Treaty of Waitangi by embedding Māori authority, accountability, and decision-making power into research data processes.
0 citationsRead paper
Rashomon Alignment
Existing approaches to measuring functional similarity between models rely on the true data distribution, making it difficult to characterize alignment of decision boundaries across the entire input space. This work proposes Rashomon Alignment (RA), a novel framework that, for the first time, evaluates functional similarity between models from a geometric perspective over the full input space without dependence on any specific data distribution. By uniformly sampling the input space and employing geometric similarity metrics, RA enables a global analysis of decision boundary alignment. Experiments across more than 90 datasets demonstrate that geometric alignment provides a complementary perspective to distribution-based alignment, and that RA effectively supports model selection, ensemble construction, and enhanced interpretability.
0 citationsRead paper
Does Role Specialization Matter for Explanation Faithfulness in Mixture-of-Experts?
This work addresses the limited faithfulness of explanations in existing Mixture-of-Experts (MoE) models, which stems primarily from high representational overlap among experts that undermines role specialization. To mitigate this issue, the authors propose a representation-level decorrelation regularization mechanism that explicitly reduces the similarity between experts’ latent representations, thereby enhancing their functional distinctiveness. The method consistently improves key faithfulness metrics—including comprehensiveness, sufficiency, and AOPC—across multiple multimodal benchmarks without compromising task performance. Furthermore, the approach demonstrates broad applicability and effectiveness within standard sparse MoE architectures, validating its potential as a general-purpose enhancement for interpretable expert-based models.
0 citationsRead paper