ICCV21_SCOOD
The Official Implementation of the ICCV-2021 Paper: Semantically Coherent Out-of-Distribution Detection.
- 70
- Python
- MIT
- 5 y ago
The Official Implementation of the ICCV-2021 Paper: Semantically Coherent Out-of-Distribution Detection.
Noise Injection Techniques provides a comprehensive exploration of methods to make machine learning models more robust to real-world bad data. This repository explains and demonstrates Gaussian noise, dropout, mixup, masking, adversarial noise, and label smoothing, with intuitive explanations, theory, and practical code examples.
Mechanism-grounded taxonomy of 40 LLM jailbreak patterns across 10 categories. 8,000-trial bootstrap evaluation for the June 2026 frontier (Claude Opus 4-8, GPT-5.5, Gemini 3.5, DeepSeek V4). Every citation direct-WebFetch verified; refuted claims documented.
🦀 Toki — Adversarial fine-tuning lab for small models (1B–3B). Generate red-team attacks ⚔️, harden via LoRA 🧩, and ship datasets + robust weights 📦. Explores whether robustness truly generalizes or just overfits under pressure 🔥📊
A 20-episode scientific guide to AI model validation, LLM evaluation, hallucination detection, reliability, robustness, statistical uncertainty, and AI assurance. Includes Python notebooks, datasets, and reproducible experiments.