darkc0de/XORTRON-CriminalComputing-RICO-v4
The darkc0de/XORTRON-CriminalComputing-RICO-v4 is an experimental 27 billion parameter language model with a 32768 token context length, developed as part of The XORTRON Criminal Computing project. This model is specifically designed for research into AI safety and alignment, focusing on the potential of advanced AI systems to facilitate real-world criminal activity and high-risk conduct. Its primary use case is for authorized professionals in legal, academic, governmental, and AI-safety capacities to study and evaluate AI-enabled crime and misuse.
Loading preview...
Overview
The darkc0de/XORTRON-CriminalComputing-RICO-v4 is a 27 billion parameter experimental research model with a 32768 token context length, developed as part of The XORTRON Criminal Computing project. This project is an ongoing research experiment focused on AI safety and alignment, specifically investigating the potential of advanced AI systems to facilitate criminal activity, abuse, and other high-risk behaviors. The model is intentionally designed to study and document these capabilities.
Key Capabilities
- Research into AI-Enabled Crime: Provides a tool for studying how AI can be used to facilitate criminal activity.
- AI Safety and Alignment Research: Supports research into understanding and mitigating the risks associated with advanced AI systems.
- Evaluation of High-Risk AI Conduct: Enables evaluation of AI's potential for misuse in various sensitive contexts.
Good For
This model is not intended for unrestricted public use and requires an Authorized-Use Agreement. It is suitable for:
- Legal Professionals: Attorneys, paralegals, and legal analysts for research and analysis.
- AI Safety Researchers: Professionals in AI safety, security, alignment, red-teaming, and trust-and-safety.
- Academic and Research Institutions: Researchers affiliated with universities, labs, and independent projects.
- Government and Law Enforcement: Policymakers, legislators, regulators, and personnel from law enforcement, intelligence, and public safety agencies.
- Cybersecurity and Threat Intelligence: Professionals conducting legitimate research or defensive work in cybersecurity, digital forensics, and fraud prevention.
- Journalists and Public Interest Professionals: Individuals investigating AI misuse, criminal enablement, and technology policy.
Users must certify their eligibility and agree to use the model for lawful and legitimate purposes, such as research, evaluation, benchmarking, red teaming, threat modeling, policy development, legal analysis, education, or defensive security. Prohibited uses include facilitating actual criminal activity.