Hope of City: A Prompt-Trace Neural Atlas of 38 Language Models

Ho Yiing Chen

doi:10.5281/zenodo.20740070

June 18, 2026Open Access

Hope of City: A Prompt-Trace Neural Atlas of 38 Language Models

View Full Paper

Ask AI

Bookmark

Share

Key Points

This research aims to create a 3D prompt-trace atlas that visualizes language models' structures and interactions.
Introduced the Hope of City as a 3D atlas for visualizing LLM structures
Included features like addressable towers, layers, and prompt flow
Developed a measurement contract separating visual release count from a measurement ledger
Presented a novel framework for understanding language model interactions in a structured manner
Enabled detailed mapping of prompt flows across different models
Facilitated cross-model comparisons with the introduction of anchors and towers

Implication

Methods paper introduces a 3D atlas of language model structures, indicating new ways to analyze them.

Abstract

A Mercury methods paper introducing Hope of City, a 3D prompt-trace atlas of 38 observed LLM structures with addressable towers, layers, prompt flow, cross-model anchors, and a measurement contract that separates visual release count from the measurement ledger. Copyright (c) 2026 Chen, Ho Yiing (norika). Licensed under Creative Commons Attribution 4.0 International (CC BY 4.0).

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Ho Yiing Chen (2026) studied this question.

synapsesocial.com/papers/6a362fbcdb0793dc1a537279 https://doi.org/10.5281/zenodo.20740070

View Full Paper

Ask AI

Bookmark

Share

Hope of City: A Prompt-Trace Neural Atlas of 38 Language Models

Key Points

Implication

Abstract

Discussion

Cite This Study

Hope of City: A Prompt-Trace Neural Atlas of 38 Language Models

Key Points

Implication

Abstract

Discussion

Cite This Study

Keep reading

More related papers

Keep reading

More related papers