About
Research & Development
AlgoTune: Pushing Language Models to Optimize Numerical Programs
TLDR: AlgoTune is a new benchmark challenging language models to optimize the runtime of 155 general-purpose numerical programs from various domains. Unlike traditional benchmarks,...
Research & Development
Boosting Code Generation: A New Approach to Self-Taught Programming and Repair
TLDR: The research introduces "Dr. Boot," a bootstrapping algorithm that teaches language models to synthesize and repair code. It addresses data scarcity and the...
Research & Development
Advancing Cardiac Motion Analysis with Synthetic CT Data Generation
TLDR: A new framework uses a Conditional Variational Autoencoder (CVAE) to create synthetic, annotated 3D cardiac CT images and motion data. This addresses the...
Research & Development
Advancing Speech Quality: A New Framework Harnesses Nonlinear Dynamics for Bandwidth Extension
TLDR: NDSI-BWE is a new adversarial framework for speech bandwidth extension that uses seven novel discriminators inspired by nonlinear dynamical systems to capture complex...
Research & Development
Efficient Interpretability for Pruned LLMs Using Transferable Autoencoders
TLDR: This research explores the transferability of Sparse Autoencoders (SAEs) for interpreting compressed large language models (LLMs). It finds that SAEs trained on original,...
Research & Development
Crafting Dynamic 3D Avatars from a Single Photo with Dream, Lift, Animate
TLDR: Dream, Lift, Animate (DLA) is a new framework that reconstructs high-fidelity, animatable 3D human avatars from a single image. It works in three...