K.NOH
Kangjun Noh
Integrated Ph.D. Student at Yonsei University.
Building reliable, robust, and trustworthy AI systems for the real world.
Research Focus
Ongoing
- Towards reliable LLM-as-a-judge — Conformal inference & Reliable LLM-as-a-judge.
- Hallucination Detection — Conformal inference & LLM response factuality.
- Bias Mitigation — Auditing protocols & counterfactual reweighting.
Previous
- Tabular LLM Prediction — Language-based modeling for OOV generalization.
- Recommender Systems — Tag-aware pipeline using TC-BERT.
Publications
-
Multi-LLM Adaptive Conformal Inference for Reliable LLM Responses ICLR 2026
-
Uncertainty-Driven Embedding Convolution ICLR 2026
-
Enhancing Causal Discovery in Chronic Diseases: The MAGIC Framework Using Multiple LLMs ISPOR Summit 2025
-
Multi-purpose Technology Commercialization Recommender System Journal of Supercomputing, 2025
-
LBC: Language-Based Classifier for Out-of-Variable Generalization NAACL 2025 (Oral)
Contact
Open to collaborations on Trustworthy AI.
- Email: shwndnjs58@yonsei.ac.kr
- GitHub: github.com/kangjun-noh
- Website: kangjun-noh.github.io