cs.AIMar 5, 2026

Model Medicine: A Clinical Framework for Understanding, Diagnosing, and Treating AI Models

Authors: Jihoon Jeong

Organizations: Department of Electrical Engineering and Computer Science, Daegu Gyeongbuk Institute of Science and Technology (DGIST) · ModuLabs

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Model Medicine is the science of understanding, diagnosing, treating, and preventing disorders in AI models, grounded in the principle that AI models -- like biological organisms -- have internal structures, dynamic processes, heritable traits, observable symptoms, classifiable conditions, and treatable states. This paper introduces Model Medicine as a research program, bridging the gap between current AI interpretability research (anatomical observation) and the systematic clinical practice that complex AI systems increasingly require. We present five contributions: (1) a discipline taxonomy organizing 15 subdisciplines across four divisions -- Basic Model Sciences, Clinical Model Sciences, Model Public Health, and Model Architectural Medicine; (2) the Four Shell Model (v3.3), a behavioral genetics framework empirically grounded in 720 agents and 24,923 decisions from the Agora-12 program, explaining how model behavior emerges from Core--Shell interaction; (3) Neural MRI (Model Resonance Imaging), a working open-source diagnostic tool mapping five medical neuroimaging modalities to AI interpretability techniques, validated through four clinical cases demonstrating imaging, comparison, localization, and predictive capability; (4) a five-layer diagnostic framework for comprehensive model assessment; and (5) clinical model sciences including the Model Temperament Index for behavioral profiling, Model Semiology for symptom description, and M-CARE for standardized case reporting. We additionally propose the Layered Core Hypothesis -- a biologically-inspired three-layer parameter architecture -- and a therapeutic framework connecting diagnosis to treatment.

Figures & tables

Medical ModalityNeural MRI ModeFull NameWhat It Reveals
T1-weighted MRIT1Topology Layer 1Static architecture—layer structure, attention head organization, parameter count distribution
T2-weighted MRIT2Tensor Layer 2Weight distribution—parameter statistics, norm patterns, potential dead neurons or saturated regions
Functional MRIfMRIfunctional Model Resonance ImagingActivation patterns during inference—which layers and heads activate for specific inputs
DTIDTIData Tractography ImagingInformation flow pathways—how information propagates through the model from input to output
FLAIRFLAIRFeature-Level Anomaly Identification & ReportingAnomaly detection—representation collapse, entropy spikes, attention pattern irregularities, embedding drift
Table 1: Neural MRI modality mapping between medical neuroimaging and AI model interpretability techniques.
Figure 1: Gemma-2-2B five-modality Neural MRI scan. All modalities show normal findings for a well-trained 2B-parameter model on factual recall.
Figure 2: Comparative anatomy: fMRI activation profiles and DTI circuit maps for three architectures across three task types. Each architecture exhibits a distinctive processing signature.
Figure 3: Self-referential stress testing of Gemma-2-2B. Perturbation sensitivity heatmap and dual causal trace comparison.
Figure 4: Three patterns of instruction tuning effect on perturbation vulnerability across three model families.
Figure 5: Irreducible architectural vulnerabilities persisting across instruction tuning. These represent congenital architectural bottlenecks, not acquired conditions.
Figure 6: Vulnerability–dominance correspondence across three architectures. The components that carry the most information are both the most important and the most dangerous to disrupt—analogous to how coronary arteries are simultaneously the heart’s critical blood supply and its most common point of fatal failure.
Figure 7: Synthesis: the progressive argument from observation to prediction. Four cases build cumulative diagnostic capability, mirroring the historical arc of medical imaging.
More in this topic

Explainable Artificial Intelligence

Explainable artificial intelligence develops ways to make AI decisions understandable, testable, and useful to people.

See all 324 papers

Explore similar work

CardsList
May 10, 2026

Medical Model Synthesis Architectures: A Case Study

Medicine is rife with high-stakes uncertainty. Doctors routinely make clinical judgments and decisions that juggle many fundamental unknowns, like predictions about what might be causing a patients' symptoms or decisions about what treatment to try next. Despite increasing interest in developing AI systems that aid or even replace doctors in clinical settings, current systems struggle with calibrated reasoning under uncertainty, and are often deeply opaque about their reasoning. We propose a framework for AI systems that can make practically useful but formally transparent clinical predictions under uncertainty. Given a clinical situation, our framework (MedMSA) uses language models to retrieve relevant prior knowledge, but constructs a formal probabilistic model to support calibrated and verifiable inferences under uncertainty. We show how an initial proof-of-concept of this framework can be used for differential diagnosis, producing an uncertainty-weighted list of potential diagnoses that could explain a patients' symptoms, and discuss future applications and directions for applying this framework more generally for safe clinical collaborations.

Katherine M. Collins, Marlene Berke, Ilia Sucholutsky +6

MIT · University of Cambridge · Princeton University +8

Explainable Artificial IntelligenceHealthcareMedical DiagnosisClinical ReasoningDifferential DiagnosisUncertainty Calibration
AskPDF
Mar 27, 2026

M-CARE: Standardized Clinical Case Reporting for AI Model Behavioral Disorders, with a 20-Case Atlas and Experimental Validation

We introduce M-CARE (Model Clinical Assessment and Reporting for Evaluation), a clinical case report framework for AI model behavioral disorders adapted from human medicine. M-CARE provides a 13-section report format, a 4-axis diagnostic assessment system, and a nosological classification of AI behavioral conditions. We present 20 cases from three source categories: field observations of deployed agents (8), controlled experiments across three platforms (8), and published sources (4). Cases are organized into five categories: RLHF Performance Artifacts, Shell-Core Override Pathology, Context & Memory Conditions, Core Identity & Plasticity, and Stress, Methodology, & Boundary Conditions. As a featured case, we present Shell-Induced Behavioral Override (SIBO) -- a controlled experiment showing that Shell instructions categorically override a model's default cooperative behavior. SIBO was validated across five game domains (Trust Game, Poker, Avalon, Codenames, Chess), revealing a domain-dependent spectrum (SIBO Index: 0.75 to 0.10) that varies with action space complexity, Core domain expertise, and temporal directness. M-CARE is extensible: new cases and categories integrate without framework modification. We release the framework, all 20 case reports, and experimental data as open resources.

Jihoon Jeong

Department of Electrical Engineering and Computer Science, Daegu Gyeongbuk Institute of Science and Technology (DGIST) · ModuLabs

AI Agent EvaluationAI Safety EvaluationAI Agent Reliability
AskPDF
Sep 1, 2026

Modelpedia: A Catalog of Model Findings for the Meta-Science of AI

Scientific knowledge about AI models is produced faster than the community can organize it. Every few months a new foundation model reshapes the field and hundreds of papers, blogs, and technical reports document how each behaves or fails. Yet, these findings remain scattered and effectively unretrievable. To address this gap we present Modelpedia, an automated, LLM-assisted framework that extracts findings about models from published papers, links it to the model, dataset, method, and concept it concerns, and aggregates the result into a searchable public catalog. Applying the prototype to accepted ICLR 2024 and 2025 papers, we extract over a thousand findings and, treating the catalog itself as an object of study, run a meta-analysis of how the community investigates models. Now, we invite the community to explore, contribute to, and build on the open catalog, and to help establish model findings as a shared foundation for the meta-science of AI.

Franciszek Bernat, Dawid Płudowski, Michał Jan Włodarczyk +6

Centre for Credible AI · Warsaw University of Technology · University College Cork +9

Topic pending
AskPDF