Home / cuhk-aim-group / neurodiscovery · GitHub

model-interpretability SKILL.md skillA

model-interpretability from cuhk-aim-group/neurodiscovery is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned

mdr add cuhk-aim-group/neurodiscovery/model-interpretability@git:20260914.cfc1dd5

Pins git:20260914.cfc1dd5 in mdr.lock. Pin the fingerprint instead.

Audit grade: A, 17 of 17 checks passed
Kind: SKILL.md skill
Repository: cuhk-aim-group/neurodiscovery
Path: skills/model-interpretability/SKILL.md
Versions recorded: 1
Latest version: git:20260914.cfc1dd5, committed Sep 14, 2026
Size: 5,248 bytes
Repository stars: 93
License: MIT

Export trained-model attention weights, target-specific Integrated Gradients or Grad-CAM, patient-level feature attributions and spatial NIfTI maps with numerical QA and required-output validation. Use for post-training explanations or attention/attribution maps, not causal effect estimation.

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# Model Interpretability Skill

Produce requested explanatory artifacts in addition to predictions and metrics.
Read [the adapter and output contract](references/adapter-contract.md) before execution.
This supports trained PyTorch models through a small explicit adapter, including
tuple-output and multimodal models. It does not make every architecture
automatically explainable or turn explanations into clinical evidence.

## Select what the user asked for

- **Input/feature attribution:** use Integrated Gradients (`ig`). Select a named
  class logit or regression output and an explicit reference baseline after the
  exact training-fitted preprocessing. Keep signed attributions and per-patient
  completeness residuals. Zero is not a neutral baseline for every modality.
- **CNN localization:** use `gradcam` with an explicitly selected 2D/3D activation
  layer and spatial input. It exports original ReLU Grad-CAM, not voxel-level
  causal effects. A zero map can be a genuine method result; inspect diagnostics.
- **Attention:** use `attention` only when the model actually produces normalized
  attention weights. Preserve layer, heads/averaging, token/ROI identity and
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint

mdr add cuhk-aim-group/neurodiscovery/model-interpretability@sha256:bad16473ede8d8d1

The line at the top of this page pins the label git:20260914.cfc1dd5, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

[![mdr](https://markdownregistry.com/badge/art_ojys32lumru6554r.svg)](https://markdownregistry.com/a/art_ojys32lumru6554r)

3 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 1 version of this file, committed on Sep 14, 2026.

version	committed	commit	size	audit
git:20260914.cfc1dd5 latest	2026-09-14	cfc1dd5	5,248 B	A	view

Audit of the latest version

A 17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.

pass: Frontmatter block present
pass: Frontmatter declares a name
pass: Frontmatter declares a description
pass: Size between 200 bytes and 200 KB (5248 bytes)
pass: No zero-width or bidi control characters
pass: No instruction hidden inside an HTML comment
pass: No link to an exfiltration or paste host
pass: No credential-shaped string
pass: No instruction to send local credentials anywhere
pass: No text hidden with inline styles
pass: No prompt-injection phrasing
pass: No curl or wget piped into a shell
pass: No recursive delete of root, home or parent
pass: No instruction to read or print local credentials
pass: No base64 blob over 200 characters
pass: No link to a raw IP address
pass: No script tag

Source

GitHub

cuhk-aim-group/neurodiscovery · 93 stars · license MIT · pushed 2026-10-23 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_ojys32lumru6554r
GET https://markdownregistry.com/api/v1/resolve?ref=cuhk-aim-group/neurodiscovery/model-interpretability
GET https://markdownregistry.com/api/v1/blob/bad16473ede8d8d17895962907207fa9977f88cb0a73b8acf752cf1633949256

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.