About
Research & Development
DINOv3 and Test-Time Training: A New Training-Free Approach for Medical Image Registration
TLDR: A new training-free medical image registration method, DINOv3+T3, uses a frozen DINOv3 encoder and optimizes deformation fields at test time in a low-dimensional...
Research & Development
TransLight: Precise Light Effect Transfer in Images
TLDR: TransLight is a new AI framework that allows users to transfer specific light effects from a reference image to a target image with...
Research & Development
Enhancing Clinical Insights from Electronic Health Records with AI: A Look at RAG vs. Long-Context Models
TLDR: This research evaluates how large language models (LLMs) can better process lengthy electronic health records (EHRs) for clinical tasks. It compares Retrieval-Augmented Generation...
Research & Development
Improving Inductive Learning in Temporal Graphs
TLDR: GTGIB is a new framework for temporal graph learning that combines Graph Structure Learning (GSL) with a Temporal Graph Information Bottleneck (TGIB). It...
Research & Development
Compressing Diffusion LLMs: A Deep Dive into Post-Training Quantization
TLDR: This research paper presents the first systematic study on applying post-training quantization (PTQ) to diffusion large language models (dLLMs). It identifies activation outliers...
Research & Development
Formalizing Geometry Challenges for AI: Introducing LeanGeo
TLDR: LeanGeo is a new framework built within the Lean 4 theorem prover that formalizes and solves complex geometry problems, including those from the...