# A Knowledge-Injection Framework for Zero-Shot Adaptation of LLMs to Delirium Prediction

arXiv: https://arxiv.org/abs/2607.20453 · HTML: https://mlsanity.com/paper/2607.20453/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 19 of its 55 references that are in the map.
- First posted on arXiv 14 May 2026.

## Builds on (most cited first, 19 of 19)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Large Language Models Can Be Easily Distracted by Irrelevant Context](https://mlsanity.com/paper/2302.00093/) (2023)
- [Quantifying Language Models' Sensitivity to Spurious Features in Prompt Design or: How I learned to start worrying about prompt formatting](https://mlsanity.com/paper/2310.11324/) (2023)
- [Retrieval Augmentation Reduces Hallucination in Conversation](https://mlsanity.com/paper/2104.07567/) (2021)
- [GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models](https://mlsanity.com/paper/2410.05229/) (2024)
- [Does Prompt Formatting Have Any Impact on LLM Performance?](https://mlsanity.com/paper/2411.10541/) (2024)
- [Almanac: Retrieval-Augmented Language Models for Clinical Medicine](https://mlsanity.com/paper/2303.01229/) (2023)
- [MedRAG: Enhancing Retrieval-augmented Generation with Knowledge Graph-Elicited Reasoning for Healthcare Copilot](https://mlsanity.com/paper/2502.04413/) (2025)
- [A Closer Look at the Limitations of Instruction Tuning](https://mlsanity.com/paper/2402.05119/) (2024)
- [MedCalc-Bench: Evaluating Large Language Models for Medical Calculations](https://mlsanity.com/paper/2406.12036/) (2024)
- [Matching Patients to Clinical Trials with Large Language Models](https://mlsanity.com/paper/2307.15051/) (2023)
- [Knowledge-aware Zero-Shot Learning: Survey and Perspective](https://mlsanity.com/paper/2103.00070/) (2021)
- [Limitations of Large Language Models in Clinical Problem-Solving Arising from Inflexible Reasoning](https://mlsanity.com/paper/2502.04381/) (2025)
- [A Systematic Review of Few-Shot Learning in Medical Imaging](https://mlsanity.com/paper/2309.11433/) (2023)
- [Characterizing LLM Abstention Behavior in Science QA with Context Perturbations](https://mlsanity.com/paper/2404.12452/) (2024)
- [Large Language Models Leverage External Knowledge to Extend Clinical Insight Beyond Language Boundaries](https://mlsanity.com/paper/2305.10163/) (2023)
- [LLMs are not Zero-Shot Reasoners for Biomedical Information Extraction](https://mlsanity.com/paper/2408.12249/) (2024)
- [Will Large Language Models Transform Clinical Prediction?](https://mlsanity.com/paper/2505.18246/) (2025)

## Abstract (arXiv)

Large language models show promise for clinical prediction, but zero-shot performance on specialized tasks is limited by incomplete domain knowledge, especially for smaller locally deployable models. We present a lightweight knowledge-injection framework for zero-shot ICU delirium prediction that augments a deterministic natural-language summary of structured electronic health record data with an external clinical knowledge report at inference time, without fine-tuning or retrieval. We evaluate LLaMA 3.1 8B and LLaMA 3.3 70B on 3,160 ICU admissions from the MIMIC IV dataset. Adding a clinically meaningful external knowledge report improves AUROC by 8.57 percentage points for the 8B model and 1.99 percentage points for the 70B model compared to no external knowledge. Relative to a GPT-5.2 frontier-model reference without external knowledge report (AUROC 68.86%), knowledge injection reduces the performance gap from 15.66 to 7.09 AUROC points for LLaMA 8B and from 5.30 to 3.31 AUROC points for LLaMA 70B. Random control reports do not improve performance and often degrade it, indicating that gains depend on clinically meaningful content rather than added prompt length alone. SHAP-based attribution further confirms that the injected knowledge is actively used during prediction. These findings suggest that inference-time knowledge injection can narrow the gap between locally deployable open-weight models and frontier closed models while preserving a practical, privacy-preserving workflow for resource-constrained clinical settings.