dataset size
The total number of examples included in a dataset, which can significantly impact a model's performance, generalization ability, and training time.
- Compress to Impress: Efficient LLM Adaptation Using a Single Gradient Step on 100 Samples
- Explaining and Mitigating Crosslingual Tokenizer Inequities
- Horizon Reduction Makes RL Scalable
- Impact of Dataset Properties on Membership Inference Vulnerability of Deep Transfer Learning
- Memory Mosaics at scale
- On the Entropy Calibration of Language Models
- Power Lines: Scaling laws for weight decay and batch size in LLM pre-training
- Private Geometric Median in Nearly-Linear Time
- The Cost of Robustness: Tighter Bounds on Parameter Complexity for Robust Memorization in ReLU Nets
- UMA: A Family of Universal Models for Atoms