• A type of machine-learning model designed for natural-language processing tasks such as language generation. LLMs are language models with many parameters, trained with self-supervised learning on a vast amount of text. The largest and most capable LLMs are generative pretrained transformers (GPTs). Modern models can be fine-tuned for specific tasks or guided by prompt engineering. These models acquire predictive power regarding syntax, semantics, and ontologies inherent in human-language corpora, but they also inherit inaccuracies and biases present in the data they are trained in. Before 2017, there were a few language models that were large as compared to capacities then available, but it was GPT-2 in 2019 that caught widespread attention, because OpenAI at first deemed it too powerful to release publicly, out of fear of malicious use. ← Wikipedia

1. Related terms: LLM-based evaluation · Retrieval-augmented generation · One-shot prompting · Zero-shot prompting
2. Previous term: Laravel
3. Next term: Large viewport
4. Random term: Constant (webglossary.info/random 🎲)