5 papers
Measuring the Depth of LLM Unlearning via Activation Patching
Jaeung Lee, Dohyun Kim, Jaemin Jo
Large language model (LLM) unlearning has emerged as a crucial post-hoc mechanism for privacy protection and AI safety, yet auditing whether target knowledge is truly erased remain…
Symetra: Visual Analytics for the Parameter Tuning Process of Symbolic Execution Engines
Donghee Hong, Minjong Kim, Sooyoung Cha +1
Symbolic execution engines such as KLEE automatically generate test cases to maximize branch coverage, but their numerous parameters make it difficult to understand the parameters'…
Suppression or Deletion: A Restoration-Based Representation-Level Analysis of Machine Unlearning
Yurim Jang, Jaeung Lee, Dohyun Kim +2
As pretrained models are increasingly shared on the web, ensuring that models can forget or delete sensitive, copyrighted, or private information upon request has become crucial. M…
Unlearning Comparator: A Visual Analytics System for Comparative Evaluation of Machine Unlearning Methods
Jaeung Lee, Suhyeon Yu, Yurim Jang +2
Machine Unlearning (MU) aims to remove target training data from a trained model so that the removed data no longer influences the model's behavior, fulfilling "right to be forgott…
GhostUMAP2: Measuring and Analyzing (r,d)-Stability of UMAP
Myeongwon Jung, Takanori Fujiwara, Jaemin Jo
Despite the widespread use of Uniform Manifold Approximation and Projection (UMAP), the impact of its stochastic optimization process on the results remains underexplored. We obser…