Selected work

Representative Papers

Semantic Alignment of Multilingual Knowledge Graphs via Contextualized Vector Projections

Dec 22, 2025arXiv.org

This work addresses the challenge of cross-lingual semantic alignment of ontology entities in multilingual knowledge graphs by proposing a context-enhanced alignment approach. It constructs semantically rich multilingual entity descriptions and fine-tunes a multilingual Transformer model to generate high-quality embeddings. Precise alignments are then achieved through cosine similarity matching combined with an adaptive threshold filtering mechanism. Evaluated on the OAEI-2022 Multifarm track, the method achieves an F1 score of 71% (recall: 78%, precision: 65%), outperforming the best baseline by 16% and significantly improving both accuracy and robustness in cross-lingual ontology alignment.

0 citationsRead paper

ModHiFi: Identifying High Fidelity predictive components for Model Modification

Nov 24, 2025

Open-weight models pose challenges for component-level modification tasks (e.g., pruning or unlearning) when training data, loss functions, and gradient information are unavailable. Method: This paper proposes ModHiFi, an efficient, label-free, gradient-free importance estimation framework. Its core innovations are: (1) the Subset Fidelity metric, which quantifies global component importance via local reconstruction behavior—a first-of-its-kind formulation; and (2) a theoretical linkage between local and global reconstruction errors grounded in Lipschitz continuity, enabling fully unsupervised, data-free importance assessment. Results: ModHiFi-P achieves 11% higher speedup over state-of-the-art pruning methods on ImageNet. ModHiFi-U enables complete, zero-fine-tuning unlearning on CIFAR-10 and demonstrates strong generalization to Swin Transformers. Collectively, ModHiFi bridges a critical gap in model interpretability and editability under minimal supervision constraints.

0 citationsRead paper

Recent publications

Latest Papers

Semantic Alignment of Multilingual Knowledge Graphs via Contextualized Vector Projections

Dec 22, 2025arXiv.org

This work addresses the challenge of cross-lingual semantic alignment of ontology entities in multilingual knowledge graphs by proposing a context-enhanced alignment approach. It constructs semantically rich multilingual entity descriptions and fine-tunes a multilingual Transformer model to generate high-quality embeddings. Precise alignments are then achieved through cosine similarity matching combined with an adaptive threshold filtering mechanism. Evaluated on the OAEI-2022 Multifarm track, the method achieves an F1 score of 71% (recall: 78%, precision: 65%), outperforming the best baseline by 16% and significantly improving both accuracy and robustness in cross-lingual ontology alignment.

0 citationsRead paper

ModHiFi: Identifying High Fidelity predictive components for Model Modification

Nov 24, 2025

Open-weight models pose challenges for component-level modification tasks (e.g., pruning or unlearning) when training data, loss functions, and gradient information are unavailable. Method: This paper proposes ModHiFi, an efficient, label-free, gradient-free importance estimation framework. Its core innovations are: (1) the Subset Fidelity metric, which quantifies global component importance via local reconstruction behavior—a first-of-its-kind formulation; and (2) a theoretical linkage between local and global reconstruction errors grounded in Lipschitz continuity, enabling fully unsupervised, data-free importance assessment. Results: ModHiFi-P achieves 11% higher speedup over state-of-the-art pruning methods on ImageNet. ModHiFi-U enables complete, zero-fine-tuning unlearning on CIFAR-10 and demonstrates strong generalization to Swin Transformers. Collectively, ModHiFi bridges a critical gap in model interpretability and editability under minimal supervision constraints.

0 citationsRead paper