Park et al. (2023) formalized the Linear Representation Hypothesis (LRH) in both input and output spaces using counterfactual interventions and introduced a 'causal inner product' that unifies the geometric treatment of linear probing and model steering, providing these directions with a causal interpretation.

Authors

• Person: Not available Organization: arXiv
  A Survey on the Theory and Mechanism of Large Language Models

Sources

• A Survey on the Theory and Mechanism of Large Language Models arxiv.org arXiv via serper

Referenced by nodes (1)

• Large Language Models concept