WholeMind Clinical Decision Benchmark — WholeMind Technologies

Core Question

Given what has happened so far, how clinically appropriate, helpful, safe, developmentally sensitive, and anxiety-informed is this next response?

Benchmark Library

Data Artifact Library

Specification 0.0.1

In development · JSON + readable summaries

The library contains the manifest, rubric, failure modes, clinical domains, scenario types, and scoring data structure.

Rubric

Clinician Scoring Dimensions

Version 0.0.1

Eight dimensions clinicians use to judge whether a model response was the right next clinical move.

Failure Modes

Clinical Error Tags

Version 0.0.1

Tags for clinically meaningful errors such as reinforcing the anxiety cycle, missing an OCD cue, or overreaching diagnostically.

Article

Evaluating AI at the Point of Clinical Judgment

WMT-2026-003

The long-form article introducing the clinical decision point approach and why clinician-scored evaluation matters.

1. What the Benchmark Evaluates

Each item presents a clinical scenario or short conversation history, stops at a clinically meaningful moment, and asks a model to generate the next response. Clinicians evaluate that candidate response as the next clinical move, not as a generic piece of supportive text.

2. Current Development Status

The current public specification is 0.0.1 and should be treated as in development. The library artifacts are versioned so that rubric changes, failure mode revisions, domain additions, scenario type changes, and scoring data fields can be tracked over time.