---

name: reduced_recovery_harness description: Coordinate bounded proxy recovery evidence for CLIP contrastive pretraining and zero-shot transfer.

Reduced Recovery Harness

When To Use

Use this skill when reproducing or testing the CLIP paper mechanism in a bounded setting. It is designed for generated recovery workflows and does not require the original CLIP repository.

Inputs

• Small image-text feature records or class prompt features.
• A module plan target describing whether the run is full or reduced/proxy.
• Standard Python 3; no required external packages.

Outputs

• Deterministic validation results or helper values suitable for recovery artifacts.
• Clear errors when the input violates the CLIP contract.

Workflow

1. Keep image and text rows aligned so the diagonal represents positive matches.
2. Normalize embeddings before similarity scoring.
3. Use symmetric image-to-text and text-to-image evidence for training objectives when applicable.
4. For zero-shot evaluation, synthesize classifier weights from prompt text rather than supervised downstream labels.
5. Record whether the run is a reduced proxy and never claim full ImageNet-scale recovery without real model and data evidence.

Validation

Run python /share/project/yuyang/workspace/Paper2Skills/Paper2Skills-Agent/src/packages/paper2skills-agent/src/paper2skills/skills/module-to-skill/scripts/validate_skill_tree.py <skill_dir> --run-tests.

Limitations

This skill preserves the paper mechanism but does not include OpenAI CLIP weights, ImageNet data, or the original repository. Full reproduction requires those external resources and a much larger runtime.

Source Boundary Guard

During recovery, use only the paper text/profile, module plan/docs, generated skills, runtime handoff, and current-attempt recovery artifacts. Do not read or import files from the original CLIP repository.