mechinterp-labeler

Manage feature labeling queues and store labels with provenance.

1|Updated Jul 9, 2024
One-click install
npx skills add https://github.com/cesaregarza/SplatNLP --skill mechinterp-labeler
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mechinterp-labeler
Source: https://github.com/cesaregarza/SplatNLP/tree/main/.claude/skills/mechinterp-labeler
Command: npx skills add https://github.com/cesaregarza/SplatNLP --skill mechinterp-labeler

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill streamlines collaborative feature labeling by providing an automated queue, centralized label storage with provenance, and tools to discover similar features for consistent labeling.

Core Features & Use Cases

  • Queue management: fetch the next feature to label, prioritize work, and track queue status.
  • Label storage & provenance: save labels with categories, notes, and source information; sync across dashboard and research states.
  • Similar features & progress: identify related features to label and monitor labeling progress across a project.

Quick Start

Typical session:

  • Get next feature: poetry run python -m splatnlp.mechinterp.cli.labeler_cli next --model ultra
  • Create a label: poetry run python -m splatnlp.mechinterp.cli.labeler_cli label --feature-id 12345 --name "Example Feature" --category tactical --source "manual"
  • Find similar: poetry run python -m splatnlp.mechinterp.cli.labeler_cli similar --feature-id 12345 --top-k 5 --model ultra

Frequently Asked Questions about mechinterp-labeler

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate feature labeling workflows with a priority queue?

Automated feature labeling workflows use a managed priority queue to fetch the next feature, store labels with provenance, and surface similar features for consistency. Interactive CLI sessions guide label creation with categories and notes while tracking overall project progress.

How do I save labels with provenance and notes for data labeling projects?

Saving labels with provenance involves creating labels with categories, notes, and source information via CLI utilities. Labels sync across dashboard and research states, ensuring centralized storage and tracking for collaborative data labeling projects.

Can I find similar features for consistent labeling across a project?

Finding similar features is supported by querying related features by feature ID and model, returning top-k matches. This surfaces related items to label, ensuring consistency across the project and streamlining the collaborative labeling process.

Do I need Python tooling and project modules to run the feature labeler CLI?

Python tooling and project modules are required, including CLI utilities and mechinterp components. Dependencies must be installed via standard package management, using commands like poetry run python to execute the labeler CLI for managing queues and labels.

What is the best way to track labeling progress across a data labeling project?

Tracking labeling progress is handled by the queue management system, which monitors queue status and overall completion across a project. Interactive CLI sessions provide updates on labeled features, ensuring consistent workflow management.