diffusion models

Replication Data for the manuscript: Adjoint-State Identifiability of Piezo-Tunable Valley Splitting in 2D Magnetic Heterostructures

Title: Replication Data for: Adjoint-State Identifiability of Piezo-Tunable Valley Splitting in 2D Magnetic Heterostructures Description: This repository contains the complete multiscale computational framework and dataset required to reproduce the findings in the manuscript "Adjoint-State Identifiability of Piezo-Tuna...

Comparative Machine Learning and Graph Neural Network Approaches for Blood-Brain Barrier Permeability Prediction

Versioned research release of the BBB permeability prediction project. This release preserves the complete source code, experimental results, documentation, figures, and reproducibility materials for the project. Contents: Phase 3: 5-fold scaffold CV classical ML (RF AUROC 0.921+/-0.024) Phase 4: Chemprop GNN ensemble...

Comparative Machine Learning and Graph Neural Network Approaches for Blood-Brain Barrier Permeability Prediction

Versioned research release of the BBB permeability prediction project. This release preserves the complete source code, experimental results, documentation, figures, and reproducibility materials for the project. Contents: Phase 3: 5-fold scaffold CV classical ML (RF AUROC 0.921+/-0.024) Phase 4: Chemprop GNN ensemble...

DTI-Guided Volumetric Spherical Harmonics Regression for Single-to-Multi-Shell dMRI Synthesis

DTI-SHNet is proposed, a single-to-multi-shell synthesis framework that operates in the real symmetric spherical harmonics (SH) coefficient domain and performs spatially aware volumetric regression and introduces a signal consistency regularization that reconstructs signals on randomly sampled canonical directions from...

Toward Bimanual Dexterous Grasp of Partially Observable Objects With Data Synthesis and Diffusion-Based Generation Approach

Data-driven bimanual dexterous grasping remains challenging due to the lack of large-scale physically valid datasets and the difficulty of generating feasible bimanual grasps from partial observations. We propose an efficient bimanual grasp synthesis pipeline that decouples bimanual grasp generation into single-hand gr...

Beyond Diffusion: A hybrid mechanistic modeling approach to accurate wax deposition prediction

Tissue Channel Modeling with Molecular Communication Simulations and Testbed

This work-in-progress (WIP) paper investigates biological tissue as a communication channel for molecular communication (MC) systems. Existing MC models typically assume simplified diffusion environments, whereas biomedical applications such as HIV reservoir detection and cancer localization require molecular signals t...

South Korean AI Policy and Economic Possibilities to 2050

South Korea’s AI for All program connects public computing support with access to services and demand for domestic models. This working paper examines what such a policy could change in the economy through 2050, combining a preserved World Development Indicators panel, official demographic projections, published potent...

AETERNA EPIGENETIC SINGULARITY: Deterministic Epigenetic Rejuvenation via 160-Day Stochastic SDE Control and Continuous Algebraic Riccati Gain Adaptation (-42.81 Years with 0.00% Oncogenic Drift)

AETERNA EPIGENETIC SINGULARITY (Version 3.0 Update) Major Milestone: 160-Day Stochastic SDE Stress-Testing, Continuous Algebraic Riccati Equations (CARE), and 768-bit Diophantine Cryptographic Blockchain Seal Author: Dimitar Prodromov (AETERNA Technologies EOOD, Bulgaria)ORCID: 0009-0004-8070-1348Authority: 0x41_45_54_...

A Physics-Informed Neural Network Atlas: Sub-Percent on Smooth Forward PDEs, Orders Worse on Inverse, High-Frequency and Real Data

Physics-informed neural networks (PINNs) are promoted as a general differential-equation solver, but the accuracy actually achieved varies by orders of magnitude across problem types, and that variation is rarely laid out in one place. This report is a method atlas: a runnable catalogue in which each of the state-of-th...

Diffusion Model-Based Video Editing: A Survey

The rapid development of diffusion models (DMs) has significantly advanced image and video applications, making "what you want is what you see" a reality. Among these, video editing has gained substantial attention and seen a swift rise in research activity, necessitating a comprehensive and systematic review of the ex...

DriftJax: A Differentiable 1-D Drift-Diffusion Solar-Cell Simulator in JAX

DriftJax is a one-dimensional, end-to-end differentiable photovoltaic cell simulator based on the drift-diffusion model and the Beer-Lambert law for optical absorption, programmed in Python using JAX. Gradients of the power conversion efficiency (and any scalar figure of merit) with respect to device design parameters...