Definition
Fine-Tuned LLM
A large language model that has been further trained on a domain-specific dataset — in Videra's case, behavioral health conversations between providers and patients — to make its outputs more accurate and contextually appropriate for that domain than a general-purpose model.
From the episode “The Behavioral Health Crisis Is a Data Problem”Hear it on the podcast1 episode
Learn the language of AI-native founders
Every week: the new terms, tools, and workflows founders are using, straight from the podcast.
Related terms in AI Fundamentals
Expert systemAn early style of AI that mimics a human expert using a knowledge base of facts and hand-written if-then rules.Explainable AI (XAI)Methods that make an AI system's decisions understandable to people, rather than leaving the model a black box.Federated learningTraining a model across many devices or servers that each keep their own data, so the raw data never has to be pooled in one place.Few-shot learningWhen a model learns a new task from just a handful of examples, often given right in the prompt.Fine-tuningTaking a pre-trained model and training it further on data for a specific task or domain so it performs better there.Foundation modelA large model trained on broad data that can be adapted to many different tasks, serving as the base for many products.Foundation Model Fine-TuningThe process of taking a pre-trained large language model (such as GPT-4, Claude, or Gemini) and continuing its training on a smaller, domain-specific dataset to specialize its behavior for a particular use case without training a model from scratch. Fine-tuning allows companies to capture the reasoning and language capabilities of a foundation model while customizing it for specific vocabularies, response formats, constraints, or domain knowledge. Loman.ai fine-tunes foundation models for the specific context of restaurant phone conversations - teaching them the nuances of taking food orders, handling 86'd items, managing reservation requests, and maintaining the appropriate tone for different restaurant types. Christian's approach: test every major foundation model, identify which performs best for each specific sub-task within the product, and fine-tune for production performance.Frontier labsThe handful of companies training the most capable models, including Anthropic, OpenAI, and Google.