Topics
254 programming, AI and security concepts, each with a one-line definition, what to learn first and where to practise it.
Showing 1–24 of 35 in AI & Machine Learning · Clear filters
AI agents
Systems where a model plans and takes actions with tools in a loop toward a goal.
15 articles · 7 practice questionsLarge language models (LLMs)
Transformer models trained on large amounts of text to predict the next token.
6 articles · 11 practice questionsLLM inference and cost
Serving a model's answers: latency, throughput, batching, GPUs and cost per token.
5 articles · 12 practice questionsModel evaluation and metrics
Measuring how well a model works: train/test splits, accuracy, precision, recall, F1 and more.
2 articles · 21 practice questionsGenerative AI
Models that produce new text, images, audio or code.
4 articlesTransformers
The attention-based neural network architecture behind modern language models.
2 articles · 6 practice questionsEmbeddings
Lists of numbers that place similar meanings close together, used for search and clustering.
2 articles · 5 practice questionsScaling
Handling more load by adding processes or machines, and what has to be shared between them.
2 articles · 5 practice questionsAI safety and alignment
Making AI systems behave helpfully and safely, for example with RLHF and safety testing.
1 article · 7 practice questionsTokens and tokenization
Splitting text into the units a language model reads and is billed by.
2 articles · 4 practice questionsAgent harness
The code around a model that manages state, tools, context, permissions and feedback.
3 articlesData leakage
Information from the test set or the future reaching training, which makes results look better than they are.
1 article · 5 practice questionsDeep learning
Machine learning with many-layered neural networks.
8 practice questionsMLOps
Practices for deploying, monitoring and retraining models in production.
7 practice questionsNeural networks
Layers of weighted connections trained by backpropagation to learn patterns.
2 articlesQuantum computing
Computing with qubits, which use superposition and entanglement for certain problems.
2 articlesReinforcement learning
Learning by trial and error from rewards.
2 articlesSmall language models (SLMs)
Compact language models that are cheaper and faster, often fine-tuned for one job.
2 articlesBias–variance trade-off
The balance between a model that is too simple and one that is too sensitive to its training data.
4 practice questionsCaching
Keeping copies of results closer to where they are needed so they are served faster.
1 article · 1 practice questionHallucination
A model stating something false with confidence.
4 practice questionsOptimization and gradient descent
Adjusting model parameters step by step to reduce a loss.
4 practice questionsAI system design
Architecting LLM applications: serving, retrieval, caching, guardrails and scaling.
1 articleArtificial general intelligence (AGI)
A hypothetical AI able to learn and perform most intellectual tasks a person can.
1 article