Across
- 3. a large collection of text and code used to train
- 5. unintended preferences or patterns in an LLM’s responses based on its training data
- 8. when an LLM generates incorrect or misleading information that sounds believable
- 9. a type of AI that creates new content, like text or images, based on prompts (2 words)
- 12. the input text given to an LLM to guide its response
- 13. request for information that you give to a system, such as a search engine or an AI model
- 14. the process of improving or polishing something to make it better, clearer, or more precise
Down
- 1. the large collection of text used to teach an LLM how to generate responses (2 words)
- 2. the process of confirming or proving that something is true, accurate, or valid
- 4. a unit of text (e.g., a word or part of a word) that an LLM reads or generates
- 6. presenting someone else's work, ideas, words, or creations as your own, without giving proper credit to the source
- 7. program that can talk with people using text or voice
- 10. to carefully examine or assess something in order to determine its value, significance, quality, or condition
- 11. to repeat a key point, question, or instruction to make it more explicit