Selected work

Representative Papers

Preserving DEG Rankings for Gene Discovery in Histology-Based Spatial Gene Expression Prediction

Sep 27, 2026

This study addresses the misalignment between reconstruction objectives and the identification of differentially expressed genes (DEGs) when predicting spatial gene expression from histology images. To this end, we propose the IDER framework, which introduces a novel differentiable IDER objective function. Without requiring predefined biological labels, this approach directly optimizes gene ranking consistency through morphological proxy contrast, effectively integrating deep learning with differentiable statistical alignment techniques. Experimental evaluations on public datasets demonstrate that the proposed framework significantly improves DEG ranking consistency and pathway enrichment overlap compared to conventional reconstruction methods. By achieving superior performance, IDER enables efficient and scalable translation from histology images to spatial transcriptomics, bridging the gap between image-based prediction and downstream biological discovery.

0 citationsRead paper

Rules or Character? Scaling Laws for AI Safety Design

Aug 13, 2026

This study investigates how AI systems should dynamically balance rule-based safety mechanisms against behavior-shaping approaches as deployment scale increases, aiming to minimize both expected harm and tail risk. We formalize this trade-off for the first time, introducing a parameter α to represent the allocation of resources between the two strategies. Incorporating factors such as filter degradation, common-mode failures, and baseline behavioral vulnerability, we evaluate the trade-off using comparative statics, a multiplicative Pareto damage model, Monte Carlo simulations, and Conditional Value-at-Risk (CVaR). Our results show that the baseline vulnerability of behavior shaping is the dominant determinant of the optimal strategy. As deployment scale grows, the optimal α* shifts modestly to substantially toward behavior shaping (Δα* = +0.01 to +0.21), reaching offsets up to 0.50 under high vulnerability; moreover, at large scales, the CVaR-optimal and expected-harm-optimal solutions converge.

0 citationsRead paper

Random Window Augmentations for Deep Learning Robustness in CT and Liver Tumor Segmentation

Oct 09, 2025

Direct application of natural-image augmentation methods to CT segmentation disregards the physical meaning of Hounsfield Unit (HU) values, leading to artifacts and poor generalizability. To address this, we propose Random Window-Width Augmentation (RWWA), a CT-specific intensity augmentation method that dynamically samples window width and level based on the empirical HU distribution—thereby preserving anatomical interpretability and physical consistency of HU values. RWWA significantly improves model robustness to low-contrast and multiphase CT images. As the first work to systematically expose the limitations of generic intensity augmentation in CT and introduce a modality-adapted enhancement strategy, RWWA achieves state-of-the-art performance across multiple multi-center liver tumor segmentation benchmarks. Notably, it yields average Dice score improvements of 2.3–4.1% on challenging cases, empirically validating the critical role of physics-aware augmentation in enhancing model generalization.

0 citationsRead paper

Recent publications

Latest Papers

Preserving DEG Rankings for Gene Discovery in Histology-Based Spatial Gene Expression Prediction

Sep 27, 2026

This study addresses the misalignment between reconstruction objectives and the identification of differentially expressed genes (DEGs) when predicting spatial gene expression from histology images. To this end, we propose the IDER framework, which introduces a novel differentiable IDER objective function. Without requiring predefined biological labels, this approach directly optimizes gene ranking consistency through morphological proxy contrast, effectively integrating deep learning with differentiable statistical alignment techniques. Experimental evaluations on public datasets demonstrate that the proposed framework significantly improves DEG ranking consistency and pathway enrichment overlap compared to conventional reconstruction methods. By achieving superior performance, IDER enables efficient and scalable translation from histology images to spatial transcriptomics, bridging the gap between image-based prediction and downstream biological discovery.

0 citationsRead paper

Rules or Character? Scaling Laws for AI Safety Design

Aug 13, 2026

This study investigates how AI systems should dynamically balance rule-based safety mechanisms against behavior-shaping approaches as deployment scale increases, aiming to minimize both expected harm and tail risk. We formalize this trade-off for the first time, introducing a parameter α to represent the allocation of resources between the two strategies. Incorporating factors such as filter degradation, common-mode failures, and baseline behavioral vulnerability, we evaluate the trade-off using comparative statics, a multiplicative Pareto damage model, Monte Carlo simulations, and Conditional Value-at-Risk (CVaR). Our results show that the baseline vulnerability of behavior shaping is the dominant determinant of the optimal strategy. As deployment scale grows, the optimal α* shifts modestly to substantially toward behavior shaping (Δα* = +0.01 to +0.21), reaching offsets up to 0.50 under high vulnerability; moreover, at large scales, the CVaR-optimal and expected-harm-optimal solutions converge.

0 citationsRead paper

Random Window Augmentations for Deep Learning Robustness in CT and Liver Tumor Segmentation

Oct 09, 2025

Direct application of natural-image augmentation methods to CT segmentation disregards the physical meaning of Hounsfield Unit (HU) values, leading to artifacts and poor generalizability. To address this, we propose Random Window-Width Augmentation (RWWA), a CT-specific intensity augmentation method that dynamically samples window width and level based on the empirical HU distribution—thereby preserving anatomical interpretability and physical consistency of HU values. RWWA significantly improves model robustness to low-contrast and multiphase CT images. As the first work to systematically expose the limitations of generic intensity augmentation in CT and introduce a modality-adapted enhancement strategy, RWWA achieves state-of-the-art performance across multiple multi-center liver tumor segmentation benchmarks. Notably, it yields average Dice score improvements of 2.3–4.1% on challenging cases, empirically validating the critical role of physics-aware augmentation in enhancing model generalization.

0 citationsRead paper