5 papers
Information-Theoretic Classifier-Free Guidance with Adaptive Schedule Optimization
Haobo Chen, Xiangxiang Xu, Yuheng Bu
Diffusion models have achieved strong performance in image, text-to-image, and video generation, where conditional generation is often controlled by classifier-free guidance (CFG).…
Style or Content? Evaluating Style Classifiers with Controlled Content Overlap
Zhuo Liu, Haozheng Du, Xiangxiang Xu +1
Style classifiers can use content cues that correlate with style labels in naturally collected data, yet we lack a systematic way to measure this reliance. We study this problem wi…
Contrastive Predictive Coding Done Right for Mutual Information Estimation
J. Jon Ryu, Pavan Yeddanapudi, Xiangxiang Xu +1
The InfoNCE objective, originally introduced for contrastive representation learning, has become a popular choice for mutual information (MI) estimation, despite its indirect conne…
Separable Computation of Information Measures
Xiangxiang Xu, Lizhong Zheng
We study a separable design for computing information measures, where the information measure is computed from learned feature representations instead of raw data. Under mild assum…
Jet Expansions of Residual Computation
Yihong Chen, Xiangxiang Xu, Yao Lu +2
We introduce a framework for expanding residual computational graphs using jets, operators that generalize truncated Taylor series. Our method provides a systematic approach to dis…