Designing Large Language Models for Specific Domains: A Case Study on Live Microbe Foods for Precision Nutrition
Bibliographic record
Abstract
Large language models (LLMs) have emerged as a powerful generative tool that can apply creativity and initiative while being efficient and user-friendly. Having been trained on enormous amounts of information, they can chat and generate data for a wide variety of problems and domains. However, this generalization results in suboptimal performance when it comes to specific and niche questions within a given domain, such as pediatric nutrition. If they are to assist domain experts, their knowledge and function must be augmented with domain-specific elements. In this work, we explore the extension of pretrained LLMs with retrieval-augment generation, few-shot examples, and tools with the goal of creating a smart child-directed nutrition assistant. Its role is to receive children's$\mathbf{2 4}$-hour dietary recalls (i.e., detailed records of all foods and beverages a person consumed in the past 24 hours) and recommend food substitutions to serve specific goals, including increasing the consumption of live microbe foods. We discuss the challenges around this development process and propose solutions that can potentially be applied in other domains and problems.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.005 | 0.025 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.004 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".