About
Research & Development
Advancing 3D Scene Understanding with GaussianCross
TLDR: GaussianCross is a novel self-supervised learning framework that leverages 3D Gaussian Splatting to enhance 3D scene understanding. It addresses challenges in 3D data...
Research & Development
Decentralized AI Training for Smarter, Private Robotic Control
TLDR: FedVLA is a novel federated learning framework for Vision-Language-Action (VLA) models in robotic manipulation. It enables distributed, privacy-preserving training by avoiding direct data...
Research & Development
Optimizing LLM Networks for Precision and Speed
TLDR: This research introduces a networked LLM system that balances information accuracy and response timeliness. It proposes routing user queries to specialized LLM clusters,...
Research & Development
Optimizing Large Language Model Efficiency with LeanK’s Smart Cache Pruning
TLDR: LeanK is a novel learning-based method that addresses the efficiency challenges of large language models (LLMs) in long-context tasks by pruning unimportant key...
Research & Development
Optimizing Diffusion Models: A Confidence-Based Prediction Strategy
TLDR: A new method called Confidence-Gated Taylor significantly accelerates Diffusion Transformers (DiTs) for visual generation. It achieves this by predicting future features only at...
Research & Development
ByteGen: A Direct Approach to Modeling Financial Market Events
TLDR: ByteGen is a novel generative model that directly processes raw byte streams of limit order book (LOB) events, eliminating the need for traditional...