model scalability
Model scalability refers to the ability of a machine learning model to maintain or improve its performance as the amount of data or complexity of the task increases. Scalable models can handle larger datasets and more complex problems without deteriorating in efficiency.
- AceReason-Nemotron: Advancing Math and Code Reasoning through Reinforcement Learning
- Discovering Important Experts for Mixture-of-Experts Models Pruning Through a Theoretical Perspective
- Fourier Analysis Network
- Handling Label Noise via Instance-Level Difficulty Modeling and Dynamic Optimization
- Language Models Can Predict Their Own Behavior
- Scaling Diffusion Transformers Efficiently via $\mu$P
- Tensor-Parallelism with Partially Synchronized Activations
- Token Bottleneck: One Token to Remember Dynamics
- UMoE: Unifying Attention and FFN with Shared Experts
- Vision Transformers with Self-Distilled Registers
- YOLOv12: Attention-Centric Real-Time Object Detectors