精简训练数据可提升大模型对事实的记牢度，减少幻觉。
Cram Less to Fit More: Training Data Pruning Improves Memorization of Facts
- 基于训练损失筛选数据，减少事实数量并均衡分布频率。
- 在半合成数据上使事实准确率达理论容量极限。
- 小模型经此方法性能媲美十倍大的模型，适合资源受限场景。
大型语言模型（LLMs）常难以在参数中有效记忆事实知识，导致幻觉频发且在知识密集型任务中表现不佳。本文从信息论角度形式化了事实记忆问题，研究训练数据分布对事实准确率的影响。当训练数据中包含的事实信息量超过模型容量时，事实准确率会低于理论上限；这一问题在事实频率呈偏态分布（如幂律分布）时更严重。为此，我们提出仅依赖训练损失的数据选择策略，旨在减少训练数据中的事实数量并平滑其频率分布。在含高熵事实的半合成数据集上，该方法显著提升事实准确率至容量极限。在标注版维基百科语料上从零预训练时，使用该方法的 GPT2-Small（110M 参数）模型相比标准训练多记住 1.3 倍实体事实，达到 10 倍大模型（1.3B 参数）在全量数据上的性能水平。
原文摘要 · Abstract (English)
Large language models (LLMs) can struggle to memorize factual knowledge in their parameters, often leading to hallucinations and poor performance on knowledge-intensive tasks. In this paper, we formalize fact memorization from an information-theoretic perspective and study how training data distributions affect fact accuracy. We show that fact accuracy is suboptimal (below the capacity limit) whenever the amount of information contained in the training data facts exceeds model capacity. This is further exacerbated when the fact frequency distribution is skewed (e.g. a power law). We propose data selection schemes based on the training loss alone that aim to limit the number of facts in the training data and flatten their frequency distribution. On semi-synthetic datasets containing high-entropy facts, our selection method effectively boosts fact accuracy to the capacity limit. When pretraining language models from scratch on an annotated Wikipedia corpus, our selection method enables a GPT2-Small model (110m parameters) to memorize 1.3X more entity facts compared to standard training, matching the performance of a 10X larger model (1.3B parameters) pretrained on the full dataset.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。