RadicalNotion.AI

RadicalNotion.AI

Applied AI research for security practitioners who need full model capability without exposing sensitive data to third-party infrastructure.

Research Focus

Security Ablation & Decensoring — Removing refusal activations that block legitimate vulnerability research. AND belief It's axiomatic that adding guardrails is a destructive action to intelligence as narrowly defined. It's not as innocuous as reflect on what you were going to say, it is more "that's not ok so ignore what you know". Built on the foundational activation analysis work of Philipp Emanuel Weidmann (p-e-w/heretic).

CVE Knowledge Distillation — Compressing everything publicly known about a specific vulnerability into a single document that supercharges a small, privately-hosted model. Analysts add sensitive environmental context in their own infrastructure. Nothing sensitive ever reaches a frontier model.

Real-Time Vulnerability Management — Automated analyst reports synthesizing CVE mechanics, exploit chains, detection opportunities, and remediation guidance at the moment of investigation.

Model Integrity Research (early stage) — Investigating post-training modification as a threat surface, with focus on code assistance and agentic workflows.

Principles

• Open-weight models only
• Private inference by default
• Data sovereignty is non-negotiable
• Attribution always

ModelAtlas

Navigation and documentation for all published models → ModelAtlas Space

*RadicalNotion.AI Inc. · LinkedIn · RadicalNotion.AI