Model Equivariance Auditor

Verifies that implemented neural network models correctly respect their intended symmetries through systematic equivariance testing, layer-wise isolation, and gradient analysis. Use when testing model equivariance, debugging symmetry bugs, verifying implementation correctness, checking if a model is actually equivariant, or diagnosing why an equivariant model isn't working.

200 stars 25 forksAI, LLMs & Agents Plugin marketplace

Install

Claude Code plugin
/plugin marketplace add lyndonkl/claude

The repo is a Claude Code plugin marketplace. Run /plugin afterwards to pick the plugin that includes this skill.

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add lyndonkl/claude --skill model-equivariance-auditor
Manual install
git clone --depth 1 https://github.com/lyndonkl/claude.git /tmp/claude
mkdir -p ~/.claude/skills
cp -r /tmp/claude/skills/model-equivariance-auditor ~/.claude/skills/model-equivariance-auditor

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Model Equivariance Auditor

Even with equivariant libraries, implementation bugs can break equivariance. A model that claims equivariance but lacks it will train poorly and give inconsistent predictions. Catching these bugs early saves significant debugging time.

## Workflow

Copy this checklist and track your progress:

```
Equivariance Audit Progress:
- [ ] Step 1: Gather model and symmetry specification
- [ ] Step 2: Run numerical equivariance tests
- [ ] Step 3: Test individual layers
- [ ] Step 4: Check gradient equivariance
- [ ] Step 5: Identify and diagnose failures
- [ ] Step 6: Document audit results
```

**Step 1: Gather model and symmetry specification**

Collect: the implemented model, the intended symmetry group, whether each output should be invariant or equivariant, the transformation functions for input and output spaces. Review the architecture specification from design phase. Clarify ambiguities with user before testing.

**Step 2: Run numerical equivariance tests**

Execute end-to-end equivariance tests using [Test Implementation](#test-implementation). For invariance: verify ||f(T(x)) - f(x)|| < ε. For equivariance: verify ||f(T(x)) - T'(f(x))|| < ε. Use multiple random inputs and transformations. Record error statistics. See [Error Interpretation](#error-interpretation) for thresholds. For ready-to-use test code, see [Test Code Templates](./resources/test-templates.md).

**Step 3: Test individual layers**
Read the full skill on GitHub

Details

Skill name
model-equivariance-auditor
Path
skills/model-equivariance-auditor/SKILL.md
Repo updated
1 September 2026

FAQ

What does the Model Equivariance Auditor skill do?

Verifies that implemented neural network models correctly respect their intended symmetries through systematic equivariance testing, layer-wise isolation, and gradient analysis. Use when testing model equivariance, debugging symmetry bugs, verifying implementation correctness, checking if a model is actually equivariant, or diagnosing why an equivariant model isn't working.

How do I install the Model Equivariance Auditor skill?

Claude Code plugin: /plugin marketplace add lyndonkl/claude. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Model Equivariance Auditor popular?

It lives in lyndonkl/claude, which has 200 GitHub stars and 25 forks.

Related skills

Abstraction Concrete Examples

200

lyndonkl/claude

Builds structured abstraction ladders that translate high-level principles into concrete, actionable examples across 3-5 levels. Bridges communication gaps, reveals hidden assumptions, and tests whether abstract ideas work in practice. Use when explaining concepts at different expertise levels, moving between abstract principles and concrete implementation, identifying edge cases by testing ideas against scenarios, designing layered documentation, decomposing complex problems into actionable steps, or bridging strategy-execution gaps.

Testing & QA25

Academic Letter Architect

200

lyndonkl/claude

Guides the creation of evidence-based academic recommendation letters, reference letters, and award nominations that combine concrete examples, meaningful comparisons, and genuine enthusiasm. Use when writing recommendation letters for students, postdocs, or colleagues, or when user mentions recommendation letter, reference, nomination, letter of support, endorsement, or needs help with strong advocacy and comparative statements.

Research & Science25

Adr Architecture

200

lyndonkl/claude

Documents significant architectural and technical decisions with full context, alternatives considered, trade-offs analyzed, and consequences understood. Creates a decision trail that helps teams understand why decisions were made. Use when choosing between technology options, making infrastructure decisions, establishing standards, migrating systems, or when user mentions ADR, architecture decision, technical decision record, or decision documentation.

Software Development25

Narrative Fidelity Audit

200

lyndonkl/claude

Runs the hard gate on invention in fact-based narrative — a bright-line refusal sweep, an interior-state provenance ladder ([A] said-so / [B] contemporaneous document / [C] behavioural inference / [D] unsourced, where any [D] blocks publication), an LLM-specific fabrication red-flag scan, a three-altitude "How do you know?" prosecution at sentence, paragraph and passage level, a subtraction-log set review, and a note on sources generated from the deviation log. Use before shipping any narrative built on real evidence — market history, incident post-mortem, biography, ML writeup, supply-chain r

AI, LLMs & Agents25

Mlb League State Reader

200

lyndonkl/claude

Parses Yahoo Fantasy Baseball league state (roster, standings, current matchup, FAAB remaining, free agents) from authenticated Yahoo team pages via Claude-in-Chrome browser automation, then grounds it against league-config.md and team-profile.md to emit a normalized league-state bundle every other agent can consume without re-scraping. Use when the coach or any downstream agent needs to read Yahoo roster, refresh team profile, pull league state, get current matchup, check FAAB remaining, list free agents, or when user mentions "what's on my roster", "who am I playing this week", "how much FAA

AI, LLMs & Agents25

Normalize Format

200

lyndonkl/claude

Normalizes a single inbox file of any supported format (plain markdown, Claude.ai JSON export, Claude Code JSONL session, Readwise markdown/CSV highlight, transcript with timestamps or speaker labels, link capture) into a clean markdown body plus partial frontmatter (id, title, source block, word_count). Handles format-specific failure modes — JSON content-block arrays, timestamp stripping, per-highlight chunking, URL-vs-commentary separation. Use when ingesting any inbox item for the substacker Librarian. Trigger keywords — normalize, convert, parse, transcript, export, JSON, JSONL, highlight

AI, LLMs & Agents25

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.