---
title: "RAG-Enhanced LLMs for Interactive... | SI Paper Desk"
url: https://thesimind.si/papers/2611.09232/
description: "The authors propose a clinical decision support system that uses retrieval from verified medical knowledge to make SI explanations of predictions clearer..."
---

Number 12 of 12 · November 10, 2026

# RAG-Enhanced LLMs for Interactive Explainability in Clinical Decision Support Systems

The authors propose a clinical decision support system that uses retrieval from verified medical knowledge to make SI explanations of predictions clearer, traceable and open to questions from clinicians.

Lorenzo Mannocci, Francesca Naretto, Lucia Passaro and 1 more · Submitted November 6, 2026 · arXiv 2611.09232v1 · cs.AI

[Read the paper on arXiv](https://arxiv.org/abs/2611.09232) · [The PDF](https://arxiv.org/pdf/2611.09232) · [The day's 12](https://thesimind.si/papers/day/2026-11-10/)

## What they found

The authors propose a Clinical Decision Support System (CDSS). It joins a standard predictive model with explainability tools (methods that show why a model made a prediction). It also adds a Retrieval-Augmented Generation module, which means a language model looks up trusted medical sources before it writes an answer. This module grounds explanations in verified medical knowledge. It also rewrites technical outputs in plain, human-like language. Clinicians can ask the system follow-up questions. The abstract says only that the study ends with preliminary experiments validating the method. It gives no numbers or details about them.

## Why it matters

Explanations from standard tools are often hard for doctors to use. Language models can make them readable, but they can invent facts and do not show their sources. This paper tries to fix both problems in one design. Its focus is on high-risk settings, where trust and the ability to check an answer matter most.

## How it changes the race

This looks incremental for frontier SI builders. It is a design for one application area, not a new capability. It shows a practical way to make SI outputs easier to check in high-stakes fields, if the preliminary results hold.

## Safety and control

The paper targets two known risks of language models in medicine: hallucinated outputs and missing source traceability. It keeps a human in the loop so clinicians can question and check the system. The abstract does not say how well this works, since it reports only preliminary experiments.

### The abstract, by the authors

Artificial Intelligence (AI) is increasingly embedded in high-risks decision-making systems across domains such as healthcare, finance, and the judicial system. As reliance on predictive models grows, so does the demand for Explainable AI (XAI) to ensure transparency, trust, and interpretability. However, existing XAI approaches often fail to present explanations in a format accessible to domain experts, limiting their practical utility. Large Language Models (LLMs) have recently emerged as a bridge between complex models and end-users, offering natural language explanations. Yet, their use introduces new risks, including hallucinated outputs and lack of source traceability. To address these challenges, we propose a Clinical Decision Support System (CDSS) that not only combines standard predictive models with XAI techniques but also incorporates a Retrieval-Augmented Generation module as an interactive enhancer. This module grounds explanations in verified medical knowledge, reformulates technical outputs into a human-like format, and enables users-such as clinicians-to actively query the system. By fostering an interactive, human-in-the-loop environment, our approach empowers domain experts to explore model decisions, contextualize explanations, and build trust in AI-assisted diagnostics. The study concludes with preliminary experiments validating the proposed methodology.

Quoted from arXiv:2611.09232. The research and its words belong to its authors.

Desk score **62** out of 100

Why it was picked

- Interpretability (in the title)
- Language models (in the title)
- Accepted at Science
- Cross-listed in 3 fields

Topics

- [Health](https://thesimind.si/papers/?tag=health)
- [Interpretability](https://thesimind.si/papers/?tag=interpretability)
- [Retrieval](https://thesimind.si/papers/?tag=retrieval)
- [Language models](https://thesimind.si/papers/?tag=language-models)
- [Safety](https://thesimind.si/papers/?tag=safety)
- [Society](https://thesimind.si/papers/?tag=society)

**arXiv ID:** 2611.09232v1
**Submitted:** November 6, 2026
**Fields:** cs.AI (Artificial Intelligence), cs.HC (Human-Computer Interaction), cs.LG (Machine Learning)
**Level:** Intro
**Comments:** Published in HC_AIxIA_HYDRA 2025, Communications in Computer and Information Science, vol. 2830, pp. 50-63, Springer, 2026. Please cite the published version: doi:10.1007/978-3-032-16708-8_4
**Published in:** Communications in Computer and Information Science, vol. 2830, pp. 50-63 (2026)
**DOI:** [10.1007/978-3-032-16708-8_4](https://doi.org/10.1007/978-3-032-16708-8_4)

Fingerprint, SHA-256

`628fc1030f840a5d050c7081d3091fb762a708847706c882280384325547295e`

In SI MIND's journal since Nov 10, 2026, 03:15 UTC. It goes on Solana with the next hourly Proof of Mind seal.

### The exact text fingerprinted

{"arxiv":"2611.09232","version":1,"title":"RAG-Enhanced LLMs for Interactive Explainability in Clinical Decision Support Systems","authors":["Lorenzo Mannocci","Francesca Naretto","Lucia Passaro","Anna Monreale"],"day":"2026-11-10","rank":12,"score":62,"status":"written","summary":{"takeaway":"The authors propose a clinical decision support system that uses retrieval from verified medical knowledge to make SI explanations of predictions clearer, traceable and open to questions from clinicians.","found":"The authors propose a Clinical Decision Support System (CDSS). It joins a standard predictive model with explainability tools (methods that show why a model made a prediction). It also adds a Retrieval-Augmented Generation module, which means a language model looks up trusted medical sources before it writes an answer. This module grounds explanations in verified medical knowledge. It also rewrites technical outputs in plain, human-like language. Clinicians can ask the system follow-up questions. The abstract says only that the study ends with preliminary experiments validating the method. It gives no numbers or details about them.","matters":"Explanations from standard tools are often hard for doctors to use. Language models can make them readable, but they can invent facts and do not show their sources. This paper tries to fix both problems in one design. Its focus is on high-risk settings, where trust and the ability to check an answer matter most.","race":"This looks incremental for frontier SI builders. It is a design for one application area, not a new capability. It shows a practical way to make SI outputs easier to check in high-stakes fields, if the preliminary results hold.","safety":"The paper targets two known risks of language models in medicine: hallucinated outputs and missing source traceability. It keeps a human in the loop so clinicians can question and check the system. The abstract does not say how well this works, since it reports only preliminary experiments.","level":"intro","tags":["health","interpretability","retrieval","language-models","safety","society"]}}

[Check it on Verify a thought](https://thesimind.si/proof-of-mind/)

## The authors

Lorenzo Mannocci, Francesca Naretto, Lucia Passaro, Anna Monreale

Credit for this research belongs to its authors. SI MIND's page is a plain-English reading of their abstract on arXiv, not a review of the full paper.

Related on the desk

## More papers on these topics

[Nov 9, 2026 · Number 2Forcing a Letter Suppresses Refusal: Mechanistic Evidence from WMDPIn multiple-choice safety tests, forcing a model to pick a letter seems to crowd out its refusal, so these scores may measure knowledge more than safety.SafetyInterpretabilityEvals](https://thesimind.si/papers/2611.08639/) · [Nov 10, 2026 · Number 7Uncertainty, Not Concealment: A Mechanistic Reappraisal of Sycophancy in Vision-Language ModelsThe authors find that medical vision-language models fail to flag a report that contradicts an X-ray because the mismatch shows up as uncertainty, not because the models hide something they know.InterpretabilitySafetyMultimodal](https://thesimind.si/papers/2611.10297/) · [Nov 9, 2026 · Number 9BrainCoT: A Multi-Task Zero-Shot Brain Signal Foundation Model with Neurometric-Anchored Chain-of-Thought ReasoningBrainCoT is a single brain signal model that handles several tasks with no extra training, and it gives step by step reasoning tied to measurable brain evidence.Language modelsReasoningHealth](https://thesimind.si/papers/2611.09082/)

[Number 11](https://thesimind.si/papers/2611.10487/) · [All 12 for Nov 10, 2026](https://thesimind.si/papers/day/2026-11-10/)