OpenSkillIndex← back to search
claude-codescience

Babysitter pymatching-decoder

Minimum-weight perfect matching decoder skill for surface code error correction

install
source · Clone the upstream repo
git clone https://github.com/a5c-ai/babysitter
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/a5c-ai/babysitter "$T" && mkdir -p ~/.claude/skills && cp -r "$T/library/specializations/domains/science/quantum-computing/skills/pymatching-decoder" ~/.claude/skills/a5c-ai-babysitter-pymatching-decoder && rm -rf "$T"
manifest: library/specializations/domains/science/quantum-computing/skills/pymatching-decoder/SKILL.md
tags
#quantum-computing#error-correction#surface-code#decoding#graph-theory#simulation
source content

PyMatching Decoder

Purpose

Provides expert guidance on minimum-weight perfect matching decoding for surface codes and other topological quantum error correction codes.

Capabilities

  • MWPM decoding for surface codes
  • Weighted edge matching
  • Detector error model processing
  • Logical error rate calculation
  • Integration with Stim simulations
  • Custom graph construction
  • Belief propagation integration
  • Parallelized decoding

Usage Guidelines

  1. Graph Construction: Build matching graph from detector error model
  2. Weight Assignment: Configure edge weights based on error probabilities
  3. Decoding Execution: Run MWPM algorithm on syndrome data
  4. Error Analysis: Calculate logical error rates from decoding results
  5. Optimization: Tune decoder parameters for specific code structures

Tools/Libraries

  • PyMatching
  • NetworkX
  • Stim
  • NumPy