-
Uncertainty Estimation Methods in Large Language Models - A Taxonomy
A systematic taxonomy of uncertainty estimation methods for Large Language Models, categorizing key literature from token-level probabilities to semantic clustering and internal state probing.
-
Reproduction and Extension - Interpretable Generative Models through Post-hoc Concept Bottlenecks (CVPR 2025)
A blog-style walkthrough of the CVPR 2025 paper on post-hoc concept bottleneck models and their interpretability.