calibration-conformal
CommunityTune calibration and conformal thresholds.
Data & Analytics#calibration#mondrian#temperature-scaling#conformal#abstention#artifact-serialization
Authorrilical
Version1.0.0
Installs0
System Documentation
What problem does it solve?
This Skill provides guidance and tooling to implement or adjust temperature scaling, split conformal, Mondrian conformal, and general abstention policies for open-world detectors, ensuring calibrated decision making and reliable abstention behavior.
Core Features & Use Cases
- Calibration workflow: fit, serialize, and apply temperature scaling and nonconformity scores for conformal prediction.
- Abstention policy tooling: support for forced decision, threshold-based, and conformal abstention in inference.
- Use Case: calibrate a detector on a held-out split and generate temperature.json and conformal.json artifacts for reuse.
Quick Start
Fit calibration on a held-out dataset, compute conformal thresholds, and serialize artifacts for reuse in evaluation and inference.
Dependency Matrix
Required Modules
None requiredComponents
Standard package💻 Claude Code Installation
Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.
Please help me install this Skill: Name: calibration-conformal Download link: https://github.com/rilical/OpenWorld-AI-Image-Detection/archive/main.zip#calibration-conformal Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
Agent Skills Search Helper
Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.