Outlier Dimensions Encode Task-Specific Knowledge
This addresses the role of outlier dimensions in LLMs for improving interpretability and efficiency in NLP tasks, though it is incremental as it builds on prior work on representation analysis.
The study investigated how fine-tuning affects outlier dimensions in LLM representations, finding that these dimensions persist post-fine-tuning and a single dimension can complete tasks with minimal error, suggesting they encode task-specific knowledge.
Representations from large language models (LLMs) are known to be dominated by a small subset of dimensions with exceedingly high variance. Previous works have argued that although ablating these outlier dimensions in LLM representations hurts downstream performance, outlier dimensions are detrimental to the representational quality of embeddings. In this study, we investigate how fine-tuning impacts outlier dimensions and show that 1) outlier dimensions that occur in pre-training persist in fine-tuned models and 2) a single outlier dimension can complete downstream tasks with a minimal error rate. Our results suggest that outlier dimensions can encode crucial task-specific knowledge and that the value of a representation in a single outlier dimension drives downstream model decisions.