Kill Webs by Collaborative & Self-organizing Agents (CSOAs)

Ying Zhao, Naval Postgraduate School, CA, USA

Charles C. Zhou, Quantum Intelligence, Inc., CA, USA

11/4/2025

1

Agenda

• Introduction
• Major Contributions
• Methods & Results
• Use Case
• Comparison
• Conclusion
• Disclaimer

2

Introduction

3

Prisoner’s dilemma: a player maximizes its own value or reaches NE, however, total welfare of the peer network is not maximized.

The QAE-QIG makes Agent cooperation is ideal if it can maximize its own value and also the peer network achieves the maximized value of total social welfare

• Collaborative and self-organizing agents (CSOA) can ingest domain specific data and self-organize to maximize quantum theoretic values (QTV) using quantum properties such as Quantum Adiabatic Evolution (QAE) and Quantum Intelligence Game (QIG)

• Kill web: potentially to improve the traditional kill chain process such as find, fix, track, target, engage, and assess (F2T2EA). The improvement is measured in the values of powerful global optimization, distributed lethality, and load balancing. We show a use case of the QAET-QIG frame to a potential application of mixed sensors, platforms, weapons, and effects.
• Finance: Identify top performers or underperformers for maximum ROI
• Predict components of functional materials, AI/ML framework for additive manufacturing defect detection, new composite materials of expected functionality with lower cost
• Biotech & Clean energy: Pinpoint efficient microbes to transform waste into energy
• Organizational Human Factors:
• College student suicide/traffic accident prevention
• DoD readiness: Addressing suicide/traffic accident risks
• Logistics and maintenance readiness

​

4

Use Cases of CSOAs

Traditional QAET

5

Major Contributions

• We show that a multi-agent quantum system converges to an equilibrium state which has a high degree of purity and coherence via a QAET-QIG evolution. The convergence results in an equilibrium and measurable property or quantum theoretic value (QTV) of the system that is optimized.
• We demonstrate the QAET-QIG framework,
• Show two theorems, one unsupervised algorithm used to score new data for quantum values
• Compare them with the traditional QAET
• We apply the QAET-QIG framework to the kill web concept that can potentially to improve the traditional kill chain process such as find, fix, track, target, engage, and assess (F2T2EA). The improvement is measured in the values of powerful global optimization, distributed lethality, and load balancing. We show a use case of the QAET-QIG frame to a potential application of mixed sensors, platforms, weapons, and effects.

6

One Agent Performing Unsupervised Learning

Multiple Agents Collaborate & Self-Organize

Kill Web Concept

7

• Objective:

​

• Look for a K that maximizes ||K |X(t)>||² for any X(t)
• QIG (H) is Kill Web (K)

​

• Kill chain is a linear sequence vulnerable to disruptions, eg, electronic warfare or anti-access strategies.
• Kill web shifts to a dynamic, mesh-like network of interconnected assets, enabling parallel operations and redundancy across domains (e.g., air, sea, cyber).
• Quantum technologies, including adiabatic optimization and game theory, could further enhance this by handling exponential complexity in real-time decision-making.

QAET-QIG integrates

​

​

8

• QAET: A variational approximation to adiabatic quantum computing, using hybrid classical-quantum methods to evolve systems toward global optima via parametrized unitaries (e.g., optimizing phase factors β and γ for coherence)
• QIG: A quantum game-theoretic model where agents (e.g., assets) interact via superposition and entanglement to reach superior equilibria, often in unsupervised or crowd-sourced scenarios.
• This framework leverages a superpositioned torus topology— a looped, multi-dimensional space allowing for entangled explorations—to converge multi-agent systems to high-purity, coherent equilibria, optimizing a quantum theoretic value (QTV) like system efficiency or decision quality.

Results

9

Example:

Spectral Norm in the Quantum Evolution

10

Kill �Web �Heatmap �Evolutions

Spectral norm K is the maximum singular value of K

Potential of QAET-QIG in Kill WEB

11

Conclusions

• We show a new QAET-QIG framework, where a repeated measurement forced by the environment with CSOAs in a QAET process.
• We also show QIG process is essential for the QAET process to converge and "collapse" to good measurement values or QTV
• The QAET-QIG is a unsupervised learning framework that can be used to score unstructured data for their values or impacts
• We apply the QAET-QIG framework to the kill web concept that potentially improve the traditional kill chain process such as find, fix, track, target, engage, and assess (F2T2EA).
• The framework ensures the kill web performance against any adversarial input is optimized. The improvement is measured in the values of global optimization, distributed lethality, and load balancing

​

​

​

12

Disclaimer

The views and conclusions contained in this document are those of the authors and should not be interpreted as representing any official policies, either expressed or implied of the U.S. Government.

13

14

Thank you!

​

Ying Zhao

Naval Postgraduate School, California, USA

yzhao@nps.edu

​

Charles C Zhou

Quantum Intelligence, Inc., California, USA

charles.zhou@quantumii.com