arXiv:2609.22099v1 Announce Type: new Abstract: Cooking is a complex process that transforms raw ingredients into delicious and nutritious dishes, yet the recipes that encode this process remain larg...
The paper investigates how well large language models (LLMs) can assess whether recipes are suitable for people with diabetes. It introduces a benchmark of 7,607 recipes, split evenly between suitable and unsuitable, and tests three prompting strategies that vary in how much diabetes dietary guidance they provide. Results show that LLMs tend to be cautious in labeling recipes as suitable, and those that can reason with dietary guidelines—particularly Mistral‑7B and Llama‑70B—perform best.
The paper examines how large language models (LLMs) can assess whether recipes are suitable for people with diabetes. It introduces a benchmark of 7,607 recipes, split evenly between suitable and unsuitable, and tests three prompting strategies that embed varying levels of diabetes dietary guidelines. Results show that LLMs tend to be cautious in labeling recipes as suitable, and those that can reason with the guidelines—particularly Mistral‑7B and Llama‑70B—perform best.
arXiv:2606. 08948v1 Announce Type: cross Abstract: Comprehensive estimation of dietary micronutrients from food images could improve clinical nutrition care, but training such models requires large multimodal datasets linking diverse foods to complete nutrient profiles.
arXiv:2607. 23273v1 Announce Type: cross Abstract: Computational nutrition needs precise ingredient data, but current databases are incomplete, inconsistent, and built for human reference rather than automated reasoning.
arXiv:2607. 08423v1 Announce Type: new Abstract: The rapid integration of Large Vision-Language Models (VLMs) into critical infrastructure promises to revolutionize personalized healthcare and dietary management.
arXiv:2606. 10120v1 Announce Type: cross Abstract: Postprandial hyperglycemia is a key risk factor for metabolic disorders; however, existing dietary guidance is often static, impractical, and insufficiently personalized, providing recommendations that are difficult to follow or not impactful.
PA-CoT (Profile-Adaptive Chain-of-Thought) is a multi-stage prompting method that explicitly analyzes user profiles before generating responses for personalized nutritional consulting. The authors introduce the QPA benchmark, comprising 200 samples with structured profiles scored on four criteria, to evaluate personalization and safety. In a comparative study against 11 other methods, PA-CoT achieves the highest average score (4.21) and leads in both personalization (4.71) and safety (4.68) with statistically significant margins.
arXiv:2608. 19875v1 Announce Type: cross Abstract: Patients often submit short, underspecified queries to healthcare chatbots that lack the patient-specific information needed to determine an appropriate response.
arXiv:2608. 10363v1 Announce Type: new Abstract: AI agents can accelerate nutrition research, but their analyses inherit the identity, semantic, and release ambiguities of the underlying data.
arXiv:2601.13880v2 Announce Type: replace Abstract: Personalized lifestyle health analysis requires long-horizon, multi-dimensional reasoning over heterogeneous lifestyle signals, and recent advances...
The paper introduces a retrieval‑augmented multi‑agent framework that automatically generates instance‑specific evaluation rubrics for medical language models. By retrieving authoritative medical evidence, decomposing it into atomic facts, and combining these with user interaction constraints, the system produces fine‑grained criteria that outperform GPT‑4o on HealthBench and LLMEval‑Med. The generated rubrics also guide response refinement, improving medical LLM output quality by 9.2%.