Safety & Alignment

Prompt injection, jailbreaks, Constitutional AI, reward hacking and mechanistic interpretability.

25concepts
408flashcards
202minutes of reading
  1. 01 Activation Steering and Representation Engineering Adding a direction to a model's residual stream at inference time changes its behaviour without any weight update, which makes concepts like refusal, sycophancy and sentiment into vectors you can add, scale, or subtract. intermediate 6m 18 cards
  2. 02 Alignment Evaluations and Frontier-Model Risk How frontier labs and governments measure dangerous capabilities, what an eval-gated release looks like, and where the regulatory regime sits in 2026. intermediate 8m 5 cards
  3. 03 Guardrail Classifiers and Content Filtering Putting a separate classifier in front of and behind the model gives you a safety layer updatable in hours rather than retrained over weeks, and it buys that agility with latency, false refusals, and a second model to keep honest. intermediate 8m 24 cards
  4. 04 Jailbreaks and Refusal Robustness How attackers reliably bypass model refusal training, why post-hoc filters are necessary but never sufficient, and how AILuminate measures what remains. intermediate 9m 5 cards
  5. 05 Prompt Injection Why LLMs cannot reliably tell instructions from data, how indirect injection weaponises retrieved content, and which partial defences are worth deploying. intermediate 8m 5 cards
  6. 06 Training Data Memorisation and Extraction Language models reproduce fragments of their training corpus verbatim, and an adversary with only API access can pull that data back out, which turns a training-set decision into a permanent disclosure risk. intermediate 8m 24 cards
  7. 07 Watermarking and Content Provenance How a statistical signal is embedded in generated text without changing its quality, why SynthID-Text scaled to production, and the robustness limits every scheme shares. intermediate 8m 15 cards