• Home
  • Analyses
    • 1 · Corpus & Ratios
    • 2 · Standardization
    • 3 · Classification
    • 4 · VAE Features
    • 5 · Shape Traversal
    • 6 · Geography
  • Interactive
    • Feature Explorer
    • Interpolation Widget
  • About
    • The Project
    • Data & Code
    • How to Cite
  1. Getting Started
  2. Shaping History at Scale
  • Getting Started
    • Shaping History at Scale
    • About the Project
    • Data & Code
  • Corpus & Shape Analysis
    • 1 · Corpus & Height-to-Width Ratios
    • 2 · Standardization Scale
  • Classification Models
    • 3 · Period Classification
  • VAE Latent Space
    • 4 · VAE Latent Space Analysis
    • 5 · Shape Traversal
  • Geographic Analysis
    • 6 · Geographic Analysis
  • Interactive Tools
    • Feature Explorer
    • Interpolation Widget
  • Reference
    • How to Cite
  1. Getting Started
  2. Shaping History at Scale

Shaping History at Scale

Computational Diplomatics and the Diachronic Evolution of Cuneiform Tablet Morphology

Shaping History at Scale

Computational Diplomatics and the Diachronic Evolution of Cuneiform Tablet Morphology over Three Millennia

Shai Gordin · Danielle Kapon Epshtain · Michael Fire Ariel University · Ben-Gurion University of the Negev · Open University of Israel

94,936 tablet images

21 historical periods

3,300 years (3500–64 BCE)

86% shape info retention

About this site

This companion website provides interactive access to all analyses published in:

Gordin, S., Kapon Epshtain, D., & Fire, M. (2025). Shaping History at Scale: Computational Diplomatics and the Diachronic Evolution of Cuneiform Tablet Morphology over Three Millennia. PNAS. DOI: [pending]

Each section corresponds to a section of the paper and a Jupyter notebook in the code repository. The two Interactive Tools pages provide live access to the VAE feature explorer and interpolation widget via Binder — no installation required.


Analyses

📐

1 · Corpus & Ratios

Height-to-width ratios across 21 periods. The three-millennium portrait→landscape shift. Log-ratio methodology.

📊

2 · Standardization Scale

Coefficient of variation as a proxy for institutional control. From Ur III (CV = 0.251) to Neo-Babylonian (CV = 0.807).

🤖

3 · Period Classification

Decision tree → CNN → ResNet50. Shape information retention fraction: 86%. Confusion matrices and per-period performance.

🧬

4 · VAE Latent Space

12-dimensional VAE bottleneck. Discrimination (X7) vs. trend (X2, X8). Diplomatic genealogies in latent space.

🎞️

5 · Shape Traversal

What does each VAE dimension physically control? Filmstrip traversals with genre control and geometric descriptor analysis.

🗺️

6 · Geography

Is the trend real or a sampling artifact? Variance decomposition, zone analysis, site-level panel, maps.


Interactive Tools

🎛️ Try it yourself — no installation required

The two interactive widgets require only a browser. Click Launch Binder to start a cloud Jupyter session with the trained VAE model pre-loaded (~2 min to start).

🎚️

Feature Explorer

Adjust individual VAE dimensions with sliders and watch the decoded tablet silhouette change in real time. Filter by period and genre.

↔︎️

Interpolation Widget

Select two tablets or two latent vectors. Visualize the morphological path between them across 11 decoded waypoints.


Data & Code

All code, data, and the trained VAE model are publicly available:

Resource Link
Code repository github.com/DigitalPasts/ShapingHistory
Dataset (VAE encodings, ratios) Zenodo DOI: 10.5281/zenodo.12787745
Trained VAE model Zenodo DOI: 10.5281/zenodo.11668219
CDLI Catalogue cdli.mpiwg-berlin.mpg.de
DigitalPasts Lab github.com/DigitalPasts
Source Code
---
title: "Shaping History at Scale"
subtitle: "Computational Diplomatics and the Diachronic Evolution of Cuneiform Tablet Morphology"
page-layout: full
toc: false
---

::: {.hero-banner}
# Shaping History at Scale

::: {.subtitle}
Computational Diplomatics and the Diachronic Evolution of Cuneiform Tablet Morphology over Three Millennia
:::

**Shai Gordin · Danielle Kapon Epshtain · Michael Fire**
*Ariel University · Ben-Gurion University of the Negev · Open University of Israel*

::: {.hero-stats}
::: {.stat-box}
[94,936]{.stat-number}
[tablet images]{.stat-label}
:::
::: {.stat-box}
[21]{.stat-number}
[historical periods]{.stat-label}
:::
::: {.stat-box}
[3,300]{.stat-number}
[years (3500–64 BCE)]{.stat-label}
:::
::: {.stat-box}
[86%]{.stat-number}
[shape info retention]{.stat-label}
:::
:::
:::

## About this site

This companion website provides interactive access to all analyses published in:

> Gordin, S., Kapon Epshtain, D., & Fire, M. (2025). **Shaping History at Scale:
> Computational Diplomatics and the Diachronic Evolution of Cuneiform Tablet
> Morphology over Three Millennia.** *PNAS*. DOI: [pending]

Each section corresponds to a section of the paper and a Jupyter notebook in the
[code repository](https://github.com/DigitalPasts/ShapingHistory). The two
**Interactive Tools** pages provide live access to the VAE feature explorer
and interpolation widget via Binder — no installation required.

---

## Analyses

::: {.analysis-grid}

::: {.analysis-card}
[📐]{.card-icon}

::: {.card-title}
[1 · Corpus & Ratios](pages/01-ratios.qmd)
:::
::: {.card-desc}
Height-to-width ratios across 21 periods. The three-millennium portrait→landscape shift. Log-ratio methodology.
:::
:::

::: {.analysis-card}
[📊]{.card-icon}

::: {.card-title}
[2 · Standardization Scale](pages/02-standardization.qmd)
:::
::: {.card-desc}
Coefficient of variation as a proxy for institutional control. From Ur III (CV = 0.251) to Neo-Babylonian (CV = 0.807).
:::
:::

::: {.analysis-card}
[🤖]{.card-icon}

::: {.card-title}
[3 · Period Classification](pages/03-classification.qmd)
:::
::: {.card-desc}
Decision tree → CNN → ResNet50. Shape information retention fraction: 86%. Confusion matrices and per-period performance.
:::
:::

::: {.analysis-card}
[🧬]{.card-icon}

::: {.card-title}
[4 · VAE Latent Space](pages/04-vae.qmd)
:::
::: {.card-desc}
12-dimensional VAE bottleneck. Discrimination (X7) vs. trend (X2, X8). Diplomatic genealogies in latent space.
:::
:::

::: {.analysis-card}
[🎞️]{.card-icon}

::: {.card-title}
[5 · Shape Traversal](pages/05-traversal.qmd)
:::
::: {.card-desc}
What does each VAE dimension physically control? Filmstrip traversals with genre control and geometric descriptor analysis.
:::
:::

::: {.analysis-card}
[🗺️]{.card-icon}

::: {.card-title}
[6 · Geography](pages/06-geography.qmd)
:::
::: {.card-desc}
Is the trend real or a sampling artifact? Variance decomposition, zone analysis, site-level panel, maps.
:::
:::

:::

---

## Interactive Tools

::: {.interactive-callout}
::: {.callout-title}
🎛️ Try it yourself — no installation required
:::
The two interactive widgets require only a browser. Click Launch Binder to start
a cloud Jupyter session with the trained VAE model pre-loaded (~2 min to start).
:::

::: {.analysis-grid}

::: {.analysis-card}
[🎚️]{.card-icon}

::: {.card-title}
[Feature Explorer](pages/interactive-explorer.qmd)
:::
::: {.card-desc}
Adjust individual VAE dimensions with sliders and watch the decoded tablet silhouette change in real time. Filter by period and genre.
:::
:::

::: {.analysis-card}
[↔️]{.card-icon}

::: {.card-title}
[Interpolation Widget](pages/interactive-interpolation.qmd)
:::
::: {.card-desc}
Select two tablets or two latent vectors. Visualize the morphological path between them across 11 decoded waypoints.
:::
:::

:::

---

## Data & Code

All code, data, and the trained VAE model are publicly available:

| Resource | Link |
|---|---|
| **Code repository** | [github.com/DigitalPasts/ShapingHistory](https://github.com/DigitalPasts/ShapingHistory) |
| **Dataset** (VAE encodings, ratios) | [Zenodo DOI: 10.5281/zenodo.12787745](https://doi.org/10.5281/zenodo.12787745) |
| **Trained VAE model** | [Zenodo DOI: 10.5281/zenodo.11668219](https://doi.org/10.5281/zenodo.11668219) |
| **CDLI Catalogue** | [cdli.mpiwg-berlin.mpg.de](https://cdli.mpiwg-berlin.mpg.de) |
| **DigitalPasts Lab** | [github.com/DigitalPasts](https://github.com/DigitalPasts) |

© 2024–2025 Gordin, Kapon Epshtain & Fire · DigitalPasts Lab

 
  • Edit this page
  • Report an issue

Built with Quarto