No system can model its own source. Empirical proof: 6 AI architectures (GPT-4, Claude, Gemini, DeepSeek, Grok, Mistral) hit the same structural wall.

The mathematical proof of AI Model Collapse via Semantic Contraction.

The first anti-AI tarpit. Rewritten in Python. Traps LLM crawlers in an infinite maze of fake pages and Markov babble.

Project on Model Collapse in LLMs – Big Data Engineering (MSc), supervised by Prof. V. Moscato, PhD G. M. Orlando and PhD D. Russo (2025)

GenProof detects model collapse risk in pre-training datasets before training begins. It combines semantic entropy, tail-density, and AI detection into a composite probability score (ICS). Built with FastAPI and scikit-learn to help ensure data quality and compliance.

Human Signal Intelligence Protocol. Behavioral data sovereignty on TON blockchain. TEE + Proof-of-Behavior + $FORE token. Fair Launch Q3 2026.

OEA: Structured Recursive Calibration for Generative Stability — empirical study of directional calibration and epistemic filtering in recursive LLM generation (4 models, 3 architectures, 82M-1.5B)

Psychohistory Prediction Engine — 138-year cycle analysis, structural pressure scoring, and falsifiable predictions for 2026-2040

Code, phase portraits, and dynamic simulations for the preprint "CREATIVE LOOP: Informational Replication Dynamics and Stability

The codebase for the project "watch me kill my language" where an LLM is iteratively finetuned on the prompts that users send, leading to a slow decline to insanity.

Experiments for my Bachelor's thesis on fine-tuning language models and analyzing model collapse on synthetic generational data.

Admissibility and parallax instability detection for AI systems

A draft algorithmic specification for estimating origin purity, AI-generated ratio, warning flags, and review readiness in AI source-preservation systems.

🛡️ Framework de défense contre le Vandalisme Cognitif et l'empoisonnement de données dans les LLMs. Analyse quantitative du révisionnisme historique, métriques de dérive morale et implémentation de preuves de réalité par hachage temporel (C2PA/Blockchain)

Governance + provenance framework to prevent AI model collapse via semantic fingerprinting, cryptographic lineage, and federated trust. DOI: https://osf.io/ufek5

Description: A draft v0.2 specification for AI origin-purity scoring, warning-flag severity, recursive synthetic risk detection, and review routing.

A Python framework for measuring and enforcing semantic stability in LLMs through Geometric Information Theory. Implements Origin Node Invariance and Causal Language Syntax (CLS) to detect and prevent model collapse.