AI Cheat Sheet 2026

The 30 highest-yield AI facts, distilled from real exam questions. Print it, save it as a PDF, or study it here — free, no sign-up.

1. Which of the following uses artificial intelligence? → All of the above
2. What does 'data drift' mean in the context of production ML models? → The statistical distribution of input features changes over time
3. What is the best course of action for a game playing issue? → Heuristic approach (Some knowledge is stored)
4. Filters "poor" answers to prevent the algorithms from responding to certain input questions. → Bias:
5. Which evaluation metric measures the overlap between generated text and human reference text using n-gram precision? → BLEU
6. What is Retrieval-Augmented Generation (RAG)? → Combining an LLM with a retrieval system to ground responses in external documents
7. In convolutional neural networks, what is the primary purpose of a pooling layer? → To reduce spatial dimensions and provide translation invariance
8. Which MLOps practice ensures that a model retrained on new data maintains or improves its performance compared to the previous version? → Model validation and comparison
9. Which technique is used to visualize which parts of an input image most influence a CNN's classification decision? → Grad-CAM (Gradient-weighted Class Activation Mapping)
10. What is a recurrent neural network (RNN) primarily designed to handle? → Sequential and time-series data with temporal dependencies
11. One of these states defines a problem in a search space. → Initial state
12. What does "AI" mean in its entirety? → Artificial Intelligence
13. An algorithm that instructs the LLM to concentrate on certain areas of the input. → Attention mechanisms:
14. What is 'semantic chunking' in the context of building RAG systems? → Dividing documents into chunks based on semantic coherence to preserve meaningful context
15. What is the significance of training data in machine learning? → It serves as input to teach the model patterns and relationships
16. In knowledge representation, how many different sorts of entities are there? → Both A and B
17. The process of adapting an LLM for a specific task or domain by training it on a smaller, relevant dataset. → What is fine tuning
18. What is the primary advantage of ensemble methods like Random Forest over a single decision tree? → They reduce variance by aggregating predictions from multiple models
19. What is the role of the 'softmax' activation function in a multi-class classification output layer? → Converts raw logits into a probability distribution summing to 1
20. What format does knowledge representation take? → IF-THEN-ELSE
21. What is 'knowledge distillation' in the context of deep learning model compression? → Training a small 'student' model to mimic the outputs of a large 'teacher' model
22. What is 'fine-tuning' a large language model (LLM)? → Continuing training of a pretrained LLM on task-specific data to adapt its behavior
23. What is dropout regularization in neural networks? → Randomly setting a fraction of neurons to zero during training to prevent overfitting
24. In deep learning, what is transfer learning? → Reusing a model pretrained on a large dataset as a starting point for a new task
25. Which of the following might artificial intelligence (AI) in healthcare accomplish? → Improved patient outcomes through personalized treatment plans
26. What does 'temperature' control in LLM text generation? → The randomness of token sampling — higher values produce more diverse outputs
27. How does artificial intelligence work? → Making a Machine intelligent
28. What possible societal effects might fear of artificial intelligence (AI) have? → Potential job displacement and unemployment
29. Which tool is commonly used for experiment tracking in ML, allowing teams to log parameters, metrics, and artifacts? → MLflow
30. In gradient descent, what does the learning rate control? → The size of the steps taken toward the minimum of the loss function