Across
- 3. A set of instructions written in code that directs the behavior of a program.
- 5. An algorithm that categorizes data into predefined classes or labels.
- 7. The initial instruction or question given to an AI model to generate a response.
- 9. The result produced by an AI model after processing input data.
- 10. An algorithm used for training neural networks by adjusting weights based on error gradients.
- 12. A visual representation created or processed by AI systems.
- 13. A type of model that learns to generate new data points from a training dataset.
- 14. A learning paradigm where agents learn by receiving rewards or penalties for their actions.
- 15. A type of machine learning where models learn patterns without labeled training data.
Down
- 1. A learning approach where a model performs tasks it hasn’t explicitly been trained on.
- 2. Relating to meaning in language, crucial for understanding context in AI models.
- 4. Technology that converts written text into spoken words.
- 6. An unobserved variable that influences observed data in statistical models.
- 8. Human languages, like English or Spanish, that AI systems aim to understand and generate.
- 11. The process of making small adjustments to improve a pre-trained model's performance.
