Independent research exploring responsible AI development through the lens of human-centered design, policy frameworks, and ethical governance.
Status: 🌱 Active research repository | Open for dialogue & collaboration
Understanding LLM capabilities, risks, and emergent behaviors • Alignment & interpretability • Human–AI interaction • Behavior evaluation & decision systems
Principles for responsible AI • Global regulatory trends • Ethical guardrails (fairness, safety, accountability, transparency) • Public-interest AI governance
Frontier AI’s implications for society • Cross-sector AI applications (health, education, governance) • Human–machine collaboration • Inclusive and equitable innovation
This space bridges technical understanding, policy thinking, and human-centered design to support safer, more ethical, and more inclusive AI futures.
- 📚 Research notes & conceptual frameworks
- 🔬 Behavioral evaluation insights from frontier LLMs
- 🌐 Evolving thought pieces & ongoing study documentation
- 🤝 Opportunities for interdisciplinary collaboration
What makes this unique:
A cross-disciplinary synthesis grounded in 20+ years of leadership across technology, policy, governance, and social innovation, including independent evaluation of multiple frontier LLMs and real-world decision environments.
Frontier LLM Behavioural Architecture Report (15 pages)
Download PDF
Note: This is an independent, self-directed study. It is not affiliated with or commissioned by any organisation, and it remains unpublished. Shared openly for learning, transparency, and collaborative research.
2-Page Summary — Frontier LLM Behavioural Architecture
Download Summary PDF
Singh, Mankaj. (2025). Frontier LLM Behavioural Architecture: A Systematic Multi-Case Evaluation Across Five Frontier Models. Zenodo. https://doi.org/10.5281/zenodo.17749013
This work is an independent, unpublished research study conducted without institutional affiliation.
I welcome engagement from:
- AI researchers & safety practitioners
- Policymakers and governance leaders
- Technologists & public-interest innovators
- Social impact and development sector professionals
If you are exploring similar questions — or wish to co-develop frameworks, compare findings, or engage in interdisciplinary inquiry — feel free to connect.
Email: mankajsingh@gmail.com
LinkedIn: https://www.linkedin.com/in/mankaj
This work is shared under the Creative Commons Attribution 4.0 International (CC BY 4.0) license.
You may use, adapt, and build upon it with proper attribution.
This repository will continue expanding with:
- Research papers & notes
- Human-centered AI evaluation methods
- Policy frameworks & governance models
- Concept papers & interdisciplinary analyses
- Cross-sector case studies