An affine map trained on The Pile transfers steering vectors from Gemma-2B to Gemma-9B and reproduces much of the large model's native steering behavior.
Towards Measuring Representational Similarity of Large Language Models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
Understanding the similarity of the numerous released large language models (LLMs) has many uses, e.g., simplifying model selection, detecting illegal model reuse, and advancing our understanding of what makes LLMs perform well. In this work, we measure the similarity of representations of a set of LLMs with 7B parameters. Our results suggest that some LLMs are substantially different from others. We identify challenges of using representational similarity measures that suggest the need of careful study of similarity scores to avoid false conclusions.
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Linear Representation Transferability Hypothesis: Leveraging Small Models to Steer Large Models
An affine map trained on The Pile transfers steering vectors from Gemma-2B to Gemma-9B and reproduces much of the large model's native steering behavior.