
c4-meta-system
Production AI defense with 7-layer protection: mathematical constraints, object-capability access, distributed O2 consensus, SVETILO ethics. First…

Production AI defense with 7-layer protection: mathematical constraints, object-capability access, distributed O2 consensus, SVETILO ethics. First…

Reconstructs legacy Windows binaries into C source by pairing Ghidra decompiler exports with local LLMs, producing compile-checked candidates and…

Collection of CVE(work) on tenserflow binary pwning it

Defensive framework that maintains a safety-focused shadow memory to detect and block prompt-injection and long-horizon threats against LLM agents…

Black-box input-stage purification defense that neutralizes backdoor attacks on object detectors via corruption, diffusion reconstruction, and DBSCAN…

Defense framework that keeps audio-language models frozen and adds a mid-layer risk gate with late-layer safety adapters to block audio jailbreaks at…

Proof-of-concept that poisons MLflow registered models via the REST API, embedding a malicious pickle to trigger RCE when the model is loaded.

Source code about machine learning and security.

reverse engineering Gemini's SynthID detection


A Python pickling decompiler and static analyzer

Invisible Hyperlinks in Physical Photographs

AI / LLM Red Team Field Manual & Consultant’s Handbook

Droidefense: Advance Android Malware Analysis Framework

Google Tensor Cracks

Practical black-box adversarial packet generation against encrypted traffic classification with minimal overhead and full packet recoverability.

Research code reproducing multi-turn LLM jailbreak experiments (FITD, MRCJ, ActorAttack, X-Teaming) from the SoK intent-oriented systematization…

An implementation of the FAIR CRQ Framework