concept suppression 1generative modeling 1generative model safety 1inference-time editing 1logit adjustment 1longitudinal ehr 1long-tailed distribution 1text-to-video generation 1video evaluation 1zero-shot adaptation 1
From the 2 of 2 linked papers with an AI index.
2 papers
cs.CV2026
Inference-Time Concept Suppression and Video-Centric Evaluation for Text-to-Video Models
Wenxuan Chen, Wenjie Feng
The paper introduces SIRUS, a training‑free, inference‑time method that suppresses specified concepts in text‑to‑video generators while preserving other content, and proposes a vid…
cs.LG2026
AdaPCLA: Adaptive Prior-Calibrated Logit Adjustment for Long-Tailed Longitudinal EHR Generation
Shuai Cui, Chen Wenxuan, Wenjie Du +3
The paper introduces AdaPCLA, a framework that improves generative models for longitudinal electronic health records by adaptively adjusting logits to better represent rare disease…