Weight Drift as a Fine-Tuning Safety Screen: Signal and Its Limits
Chansu Han (National Institute of Information and Communications Technology)*; Takeshi Takahashi (National Institute of Information and Communications Technology)
14:20–14:40Paper ID: 74Full paper
HYSOPT-ADAM: Blockwise Hysteresis-Gated Optimization for Reinforcement Learning in Autonomous Cyber Defense
Mainak Basak (Hankyong National University); Ji-Yun Lee (Hankyong National University); Geon-Yun Shin (Hankyong National University)*
14:40–15:00Paper ID: 216Full paper
Privacy-Preserving Credit Card Fraud Data via Quantile Transformation and Noisy Histograms
Jaecheol Kim (Dongguk University); Mijeong Kim (Dongguk University); Yunsik Son (Dongguk University)*
15:00–15:15Paper ID: 102Short paper
From CVEs to Countermeasures: Benchmarking D3FEND Selection on Attack-Defense Scenarios
Shincheol Lee (National Institute of Information and Communications Technology)*; Chansu Han (National Institute of Information and Communications Technology); Insik Choi (Sejong University); Takeshi Takahashi (National Institute of Information and Communications Technology)
15:15–15:30Paper ID: 306Short paper
Development of a Technique for Non-Developers to Automatically Reflect Real-Time Changes in Specialized Domains
Jung-Sook Kim (Kimpo University)*
15:30–15:45Paper ID: 1002Late Breaking
GORIN: A Research Instrument for Studying Recognition of Covert Manipulation Tactics by Generative AI Systems
Ryan Young (Toyo University)*; Takayuki Fujimoto (Toyo University)
Presentations finish at 15:45. The scheduled session ends at 16:00.
Full paper: 15 minutes presentation + 5 minutes Q&A. Short, abstract-only and late breaking papers: 10 minutes presentation + 5 minutes Q&A. Speakers should bring their own laptop; presentation PCs are not provided.