How to inject knowledge efficiently? Knowledge Infusion Scaling Law for Pre-training Large Language Models
This addresses the challenge of balancing domain specialization and knowledge retention in LLMs, offering a practical solution for efficient knowledge infusion, though it is incremental as it builds on existing scaling law concepts.
The paper tackles the problem of catastrophic forgetting in large language models when infusing domain-specific knowledge during pre-training, proposing a scaling law that predicts optimal infusion amounts to improve downstream performance, validated across different model sizes with systematic experiments.
Large language models (LLMs) have attracted significant attention due to their impressive general capabilities across diverse downstream tasks. However, without domain-specific optimization, they often underperform on specialized knowledge benchmarks and even produce hallucination. Recent studies show that strategically infusing domain knowledge during pretraining can substantially improve downstream performance. A critical challenge lies in balancing this infusion trade-off: injecting too little domain-specific data yields insufficient specialization, whereas excessive infusion triggers catastrophic forgetting of previously acquired knowledge. In this work, we focus on the phenomenon of memory collapse induced by over-infusion. Through systematic experiments, we make two key observations, i.e. 1) Critical collapse point: each model exhibits a threshold beyond which its knowledge retention capabilities sharply degrade. 2) Scale correlation: these collapse points scale consistently with the model's size. Building on these insights, we propose a knowledge infusion scaling law that predicts the optimal amount of domain knowledge to inject into large LLMs by analyzing their smaller counterparts. Extensive experiments across different model sizes and pertaining token budgets validate both the effectiveness and generalizability of our scaling law.