How Do Large Language Models Acquire Factual Knowledge During Pretraining?

• URL: http://arxiv.org/abs/2406.11813v3
• Date: Tue, 12 Nov 2024 16:38:37 GMT
• Title: How Do Large Language Models Acquire Factual Knowledge During Pretraining?
• Authors: Hoyeon Chang, Jinho Park, Seonghyeon Ye, Sohee Yang, Youngkyung Seo, Du-Seong Chang, Minjoon Seo,
• Abstract summary: We study how large language models (LLMs) acquire factual knowledge during pretraining. Findings reveal several important insights into the dynamics of factual knowledge acquisition during pretraining.
• Score: 36.59608982935844
• License:
• Abstract: Despite the recent observation that large language models (LLMs) can store substantial factual knowledge, there is a limited understanding of the mechanisms of how they acquire factual knowledge through pretraining. This work addresses this gap by studying how LLMs acquire factual knowledge during pretraining. The findings reveal several important insights into the dynamics of factual knowledge acquisition during pretraining. First, counterintuitively, we observe that pretraining on more data shows no significant improvement in the model's capability to acquire and maintain factual knowledge. Next, there is a power-law relationship between training steps and forgetting of memorization and generalization of factual knowledge, and LLMs trained with duplicated training data exhibit faster forgetting. Third, training LLMs with larger batch sizes can enhance the models' robustness to forgetting. Overall, our observations suggest that factual knowledge acquisition in LLM pretraining occurs by progressively increasing the probability of factual knowledge presented in the pretraining data at each step. However, this increase is diluted by subsequent forgetting. Based on this interpretation, we demonstrate that we can provide plausible explanations for recently observed behaviors of LLMs, such as the poor performance of LLMs on long-tail knowledge and the benefits of deduplicating the pretraining corpus.

Related papers

• Dynamic Uncertainty Ranking: Enhancing In-Context Learning for Long-Tail Knowledge in LLMs [50.29035873837]
  Large language models (LLMs) can learn vast amounts of knowledge from diverse domains during pre-training. Long-tail knowledge from specialized domains is often scarce and underrepresented, rarely appearing in the models' memorization. We propose a reinforcement learning-based dynamic uncertainty ranking method for ICL that accounts for the varying impact of each retrieved sample on LLM predictions.
  arXiv  Detail & Related papers  (2024-10-31T03:42:17Z)
• Formality is Favored: Unraveling the Learning Preferences of Large Language Models on Data with Conflicting Knowledge [55.65162959527848]
  Large language models have shown excellent performance on many knowledge-intensive tasks. However, pretraining data tends to contain misleading and even conflicting information. This study systematically analyze LLMs' learning preferences for data with conflicting knowledge.
  arXiv  Detail & Related papers  (2024-10-07T06:49:41Z)
• On the Role of Long-tail Knowledge in Retrieval Augmented Large Language Models [33.08049246893537]
  Retrieval augmented generation (RAG) exhibits outstanding performance in promoting the knowledge capabilities of large language models (LLMs) We propose a simple but effective long-tail knowledge detection method for LLMs. Our method achieves over 4x speedup in average inference time and consistent performance improvement in downstream tasks.
  arXiv  Detail & Related papers  (2024-06-24T07:17:59Z)
• Source-Aware Training Enables Knowledge Attribution in Language Models [81.13048060332775]
  Intrinsic source citation can enhance transparency, interpretability, and verifiability. Our training recipe can enable faithful attribution to the pretraining data without a substantial impact on the model's perplexity.
  arXiv  Detail & Related papers  (2024-04-01T09:39:38Z)
• An Empirical Study of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning [70.48605869773814]
  Catastrophic forgetting (CF) is a phenomenon that occurs in machine learning when a model forgets previously learned information while acquiring new knowledge. This study empirically evaluates the forgetting phenomenon in large language models (LLMs) during continual instruction tuning.
  arXiv  Detail & Related papers  (2023-08-17T02:53:23Z)
• Knowledge Inheritance for Pre-trained Language Models [57.51305807391381]
  We introduce a novel pre-training framework named "knowledge inheritance" (KI) KI combines both self-learning and teacher-guided learning to efficiently train larger PLMs. We show that KI can well support lifelong learning and knowledge transfer.
  arXiv  Detail & Related papers  (2021-05-28T14:43:26Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.