Understanding the Effect of Noise in LLM Training Data with Algorithmic Chains of Thought

During both pretraining and fine-tuning, Large Language Models (\textbf{LLMs}) are trained on trillions of tokens of text of widely varying quality. Both phases of training typically involve heuristically filtering out ``low-quality'' or \textit{noisy} training samples, yet little is known quantitatively about how the type or intensity of noise affects downstream performance. In this work, we study how noise in chain of thought (\textbf{CoT}) impacts task performance in the highly-controlled setting of algorithmically solvable tasks. First, we develop the Traced Integer …

cs.lg data filtering fine-tuning intensity language language models large language large language models llm llms low noise pretraining quality samples text thought tokens training training data type understanding