README.md

---
license: cc-by-4.0
language:
- de
tags:
- embeddings
- curriculum
- education
- german
- sentence-transformers
---

# German Curriculum Concept Embeddings

Sentence embeddings for three focus concepts from the German school curriculum analysis project.

**Model:** `paraphrase-multilingual-mpnet-base-v2`  
**Generated:** 2026-05-10T09:50:17  

## Structure

```
concept/
  embeddings.npy          # float32 (N, 768) L2-normalised
  metadata.parquet        # one row per excerpt, all CSV columns
  metadata_preview.json   # schema + first 5 rows
```

## Concepts

### `mensch`  — 4,852 excerpts · dim=768
### `verhalten`  — 1,155 excerpts · dim=768
### `evolution`  — 1,085 excerpts · dim=768

## Usage

```python
import numpy as np
import pandas as pd
from huggingface_hub import hf_hub_download

concept = 'evolution'

emb_path  = hf_hub_download(
    repo_id="deirdosh/curriculum_embeddings",
    repo_type="dataset",
    filename=f"{concept}/embeddings.npy")

meta_path = hf_hub_download(
    repo_id="deirdosh/curriculum_embeddings",
    repo_type="dataset",
    filename=f"{concept}/metadata.parquet")

embs = np.load(emb_path)          # (N, 768)
meta = pd.read_parquet(meta_path)  # N rows
```

## Source

Generated by the [Concept Atlas Space](https://huggingface.co/spaces/deirdosh/curriculum_analysis_german).