Research ArticlePublished: 2026-03-10Open Access
Eliciting Latent Knowledge: Probing Internal Representations of Large Language Models
Dr. Evelyn Vance • Dr. Charles Zhang
Digital Object Identifier
Abstract
This paper introduces a probing framework to extract factual beliefs and latent reasoning paths from LLMs, addressing hallucination and alignment.
Subject Keywords
How to Cite
Dr. Evelyn Vance, Dr. Charles Zhang (2026). Eliciting Latent Knowledge: Probing Internal Representations of Large Language Models. Journal of Ai Safety Governance, DOI: 10.12345/ais.2026.001.
Published In
Ai Safety Governance
A peer-reviewed open access journal published by Scholarly Open.
Visit Journal HomepageLicensing & Rights
Creative Commons Attribution (CC BY 4.0): Anyone may read, download, copy, distribute, print, or link to this article.
Copyright remains with the authors under open access policy.
Learn about Open Access Policy