Does AI Dream of Electric Sheep? Testing Distance-Engagement Oscillation in LLM Perceptual Reframing

Read full text Summarise with VIRAT
Authors: Gökhan Gökmen
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Correction, October 2026. This working paper reports results as coming from three language models: Llama 3.3 70B, Qwen3 32B and Llama 4 Scout 17B. This is incorrect. Because of a configuration error in the experiment code, the model setting was never applied. Every response, including the self-scores, was generated by Llama 3.3 70B. The study is therefore a single-model study, run as three independent batches of the same design. The cross-model claims are withdrawn; the results labelled Llama 70B stand. See the correction notice on page 1 of the PDF.

Fields of Study & Subjects

Topic Modeling Multimodal Machine Learning Applications Philosophy and Theoretical Science

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
27 March 2026
Language
EN