yzj623/pplab/adaptive-unet-spec

Claude Code and agent skill by yzj623/pplab

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Validate paper-critical guided diffusion UNet architecture flags before claiming architecture-faithful recovery or sampling.

Install

npx vskill@latest install yzj623/pplab/adaptive-unet-spec

SKILL.md

The same SKILL.md was published earlier as yzj623/pplab/adaptive-unet-spec.

Adaptive UNet Architecture Specification

Use this skill when checking whether a guided-diffusion configuration preserves the architecture choices emphasized in Diffusion Models Beat GANs on Image Synthesis. It is useful for module planning, recovery manifests, and ablation/stress checks.

Inputs

  • Resolution, base channels, residual block count, and attention resolutions.
  • Flags for use_scale_shift_norm, resblock_updown, learn_sigma, and class conditioning.
  • A declaration of whether the run is full, reduced, or proxy.

Outputs

  • Normalized spec dictionary.
  • ok boolean for paper-like architecture claims.
  • Warnings describing omissions that are acceptable only for declared proxies.

Workflow

  1. Normalize CLI or JSON fields into a consistent spec.
  2. Require scale-shift adaptive normalization for paper-like guided diffusion.
  3. Require attention metadata for ImageNet-style conditional synthesis.
  4. Require residual up/downsampling and learned sigma for full high-resolution claims.
  5. If a reduced proxy does not instantiate a UNet, mark the module as cross-checked instead of fully executed.

Validation

Run python scripts/architecture_spec.py --example paper128. The tests check a passing paper-like spec and failing degraded specs.

Limitations

This skill does not build neural layers or load checkpoints. It prevents architecture metadata drift and records which architecture components were omitted in bounded recovery.

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
5

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/yzj623/pplab/adaptive-unet-spec/badge)](https://verified-skill.com/skills/yzj623/pplab/adaptive-unet-spec)

Something wrong with this skill or its page? Tell us.