2 ms·
How is it different from genetic algorithms that mutate the syntax tree until the target output is achieved?
by behnamoh 2y ago
How is it different from genetic algorithms that mutate the syntax tree until the target output is achieved?
- Karellen 2y agoIt's different in the same way that using an LLM instead of a traditional Markov chain is a different way of generating text. You're still predicting the next word at a time to hopefully end up with plausible sentences/paragraphs, but the difference is in how you model the training dataset, and how you use that model to make each next choice in your live application.