Dataset distillation creates a tiny synthetic training set that, when used with a fixed network initialization, produces models whose performance approximates that of models trained on the full original dataset.
Title resolution pending
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
representative citing papers
o1-like models overthink easy tasks; self-training reduces compute use without accuracy loss on GSM8K, MATH500, GPQA, and AIME.
VICReg prevents collapse in self-supervised image embeddings via explicit variance, invariance, and covariance regularization and matches state-of-the-art downstream performance.
citing papers explorer
-
Dataset Distillation
Dataset distillation creates a tiny synthetic training set that, when used with a fixed network initialization, produces models whose performance approximates that of models trained on the full original dataset.
-
Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs
o1-like models overthink easy tasks; self-training reduces compute use without accuracy loss on GSM8K, MATH500, GPQA, and AIME.