AI Lab / Break before believing / Challenge Board8 objects

AI Lab / failure zoo

Break before believing

Prompt injection, tokenizer traps, hallucination checks, and eval cases.

Browse this shelf

Scoped search, sorting, and page links.

Showing 8 of 8 objects in this shelf.

  1. AI Lab / AI Failures And Evals / Essays Agent Failure Zoo

    A list of common AI-agent failure modes and how to test for them.

    Beginner series seed Active AI Failures And Evals
  2. AI Lab / AI Failures And Evals / Resources Break The LLM Challenge

    Small challenges that teach how AI systems fail.

    Beginner challenge shelf Seed AI Failures And Evals
  3. AI Lab / AI Failures And Evals / Puzzles Prompt Injection Museum Guard

    An AI reads a web page. The page tries to trick it. What should the AI ignore?

    Intermediate 8 min Active AI Failures And Evals
  4. AI Lab / AI Mechanics / Puzzles Tokenizer Trap

    LLMs read chunks of text called tokens. That makes some simple-looking counting tasks harder.

    Beginner 6 min Active AI Mechanics
  5. AI Lab / AI Learning Tracks / Tracks LLM Builder: Zero to Shipped Agent

    Learn how to build a small AI app that works outside a demo.

    Beginner 6 weeks Active AI Learning Tracks
  6. AI Lab / AI Source Boundaries / Resources RAG Source Boundary Checklist

    A checklist for building AI search or RAG systems that know which sources they are allowed to trust.

    Beginner 20 min Active AI Source Boundaries
  7. AI Lab / AI Failures And Evals / Essays Evals Before Agents

    Before building an AI agent, write the checks that would prove it is helping and not taking confusing action.

    Beginner 10 min Active AI Failures And Evals
  8. Work And Systems / Work Projects / Projects AI Student Helper System

    A planned AI helper for students that focuses on verified sources, deadlines, and practical next steps.

    Intermediate project seed Seed Work Projects