Phase 2: Control & Containment · ~45 minutes · Python (stdlib · category-tagged classifier simulator)

Llama Guard and Input/Output Classification

Llama Guard 3 (Meta, Llama-3.1-8B base, fine-tuned for content safety) classifies both LLM inputs and outputs against an MLCommons 13-hazard taxonomy across 8 languages.

Hiring signal: Can operate llama guard and input/output classification in production

Introduction

A 1B-INT4 quantized variant runs at over 30 tokens/sec on mobile CPUs. Llama Guard 4 is multimodal (image + text), expands to the S1–S14 category set (including S14 Code Interpreter Abuse), and is a drop-in replacement for Llama Guard 3 8B/11B. NVIDIA NeMo Guardrails v0.20.0 (January 2026) adds Colang dialog-flow rails on top of input and output rails. The honest note: "Bypassing Prompt Injection and Jailbreak Detection in LLM Guardrails" (Huang et al., arXiv:2504.11168) showed Emoji Smuggling hit 100% attack success rate on six prominent guard systems; NeMo Guard Detect recorded 72.54% ASR on jailbreaks. Classifiers are a layer, not a solution.

Type: Learn Languages: Python (stdlib, category-tagged classifier simulator) Prerequisites: Phase 15 · 10 (Permission modes), Phase 15 · 17 (Constitution) Time: ~45 minutes

Objective

Learning objectives

Unlock the full lesson

You've read the first 2 sections. The rest of this lesson covers The Problem, The Concept, Build, Check Yourself, Key Terms & Next — plus a hands-on lab, quiz, and project artifact.

Create a free account to unlock Phase 0 and Phase 1 of every course — no credit card.

Key terms in this lesson

Guardrails