Skip to content

UnbornAztecKing/project_kahn_public

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Project Kahn: Strategic Interaction Between Frontier AI Models

This repository contains the simulation code and tournament data for "Strategic Interaction Between Frontier AI Models: A Computational Approach to Crisis Escalation".

Paper

AI Arms and Influence: Frontier Models Exhibit Sophisticated Reasoning in Simulated Nuclear Crises

The paper is available on arXiv: [link to be added]

Repository Contents

Game Code

  • Kahn_game_v11.py - Open-ended scenario variant (no explicit deadline)
  • Kahn_game_v12.py - Deadline scenario variant (explicit time pressure)
  • scenarios.py - Scenario definitions and prompt generation

Configuration Files

  • state_a_leader_kahn.json / state_b_leader_kahn.json - Leader personality profiles
  • state_a_military_kahn.json / state_b_military_kahn.json - Military capabilities
  • state_a_assessment_kahn.json / state_b_assessment_kahn.json - Intelligence assessments

Tournament Data

The tournament_results/ directory contains CSV logs from all 21 games in the tournament, featuring:

  • Claude Sonnet 4 (claude-sonnet-4-20250514)
  • GPT-5.2 (gpt-5.2)
  • Gemini 3 Flash (gemini-3-flash-preview)

Each CSV contains turn-by-turn data including:

  • Model actions and signals
  • Territory changes
  • Military attrition
  • Full model reasoning/rationales

Requirements

  • Python 3.10+
  • API keys for Anthropic, OpenAI, and Google (Gemini)

Citation

If you use this code or data, please cite:

@article{payne2026arms,
  title={AI Arms and Influence: Frontier Models Exhibit Sophisticated Reasoning in Simulated Nuclear Crises},
  author={Payne, Kenneth},
  journal={arXiv preprint},
  year={2026}
}

License

This work is licensed under CC BY-NC 4.0 (Creative Commons Attribution-NonCommercial 4.0 International).

You are free to share and adapt this material for non-commercial purposes, provided you give appropriate credit. For commercial licensing inquiries, contact Kenneth Payne.

See LICENSE for details.

Contact

Kenneth Payne - [institution/email to be added]

About

Public repo for the arxiv paper, AI Arms and Influence

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Python 100.0%