The authors of the paper 'Do LLMs Build World Representations? Probing Through the Lens of Causal Mediation' (arXiv:2303.00333) introduced CALM, a general analysis framework designed to study the linguistic competence of Large Language Models using causal probing.

Authors

• Person: Not available Organization: arXiv
  [PDF] arXiv:2303.00333v5 [cs.CL] 20 Dec 2024

Sources

• [PDF] arXiv:2303.00333v5 [cs.CL] 20 Dec 2024 arxiv.org arXiv via serper

Referenced by nodes (1)

• Large Language Models concept