Update README.md

README.md

@@ -1,93 +1,93 @@
----
-language:
-- zh
-- en
-tags:
-- embeddings
-- retrieval
-- transformer-free
-- safetensors
-- edge-ai
-license: mit
----
-
-# CleanOwl-0.1
-
-**I hate AI-SLOP SO I MADE THIS.**
-
-CleanOwl is a lightweight human-likeness scoring engine.
-
-It detects whether a sentence feels like a natural human message or AI-generated content, using:
-
-- token distribution irregularity
-- semantic continuity
-- punctuation behavior
-
-No transformer. No fine-tuning. Pure statistical signals.
-
-## Score Interpretation
-
-| Score | Meaning |
-|------|--------|
-| < 60 | Likely AI-generated / formal text |
-| 60–75 | Mixed / ambiguous |
-| > 75 | Likely human-like message |
-
-Note: This is not a classifier, but a heuristic scoring system.
-
-## Limitations
-
-- Short sentences may be misclassified
-- Highly polished human writing (e.g. essays) may look like AI
-- AI can sometimes mimic human irregularity
-
-This is a lightweight detector, not a definitive AI classifier.
-
-## Quickstart
-
-```bash
-git clone https://huggingface.co/WangKaiLin/CleanOwl-0.1
-cd CleanOwl-0.1
-
-pip install numpy safetensors
-
-python ai_score.py
-
-# or embedding entry
-python quickstart.py
-```
-
-## Example:
-
-
-```bash
-請輸入文字：先思考：在 AI 時代，什麼樣的人才不會被取代？我的答案是：具備溝通能力的人、擁有韌性的人，以及始終願意站在第一線的人。
-
-human score: 47.13
-label: ai_slop_like
-
-請輸入文字：身為專業的肥宅 都會把脂肪放在身上
-
-human score: 76.88
-label: maybe_human_like
-```
-
-## Repository Structure
-
-```bash
-CleanOwl-0.1/
-├─ ai_score.py          # human score / ai slop score
-├─ quickstart.py        # demo CLI
-├─ engine.py            # PipeOwl tokenizer + emb loader
-├─ pipeowl.safetensors  # embeddings + delta_field
-├─ tokenizer.json
-├─ ptt.npy              # style field
-├─ config.json
-├─ README.md
-├─ example.md
-└─ LICENSE
-```
-
-## LICENSE
-
+---
+language:
+- zh
+- en
+tags:
+- embeddings
+- retrieval
+- transformer-free
+- safetensors
+- edge-ai
+license: mit
+---
+
+# CleanOwl-0.1
+
+**I HATE AI-SLOP SO I MADE THIS.**
+
+CleanOwl is a lightweight human-likeness scoring engine.
+
+It detects whether a sentence feels like a natural human message or AI-generated content, using:
+
+- token distribution irregularity
+- semantic continuity
+- punctuation behavior
+
+No transformer. No fine-tuning. Pure statistical signals.
+
+## Score Interpretation
+
+| Score | Meaning |
+|------|--------|
+| < 60 | Likely AI-generated / formal text |
+| 60–75 | Mixed / ambiguous |
+| > 75 | Likely human-like message |
+
+Note: This is not a classifier, but a heuristic scoring system.
+
+## Limitations
+
+- Short sentences may be misclassified
+- Highly polished human writing (e.g. essays) may look like AI
+- AI can sometimes mimic human irregularity
+
+This is a lightweight detector, not a definitive AI classifier.
+
+## Quickstart
+
+```bash
+git clone https://huggingface.co/WangKaiLin/CleanOwl-0.1
+cd CleanOwl-0.1
+
+pip install numpy safetensors
+
+python ai_score.py
+
+# or embedding entry
+python quickstart.py
+```
+
+## Example:
+
+
+```bash
+請輸入文字：先思考：在 AI 時代，什麼樣的人才不會被取代？我的答案是：具備溝通能力的人、擁有韌性的人，以及始終願意站在第一線的人。
+
+human score: 47.13
+label: ai_slop_like
+
+請輸入文字：身為專業的肥宅 都會把脂肪放在身上
+
+human score: 76.88
+label: maybe_human_like
+```
+
+## Repository Structure
+
+```bash
+CleanOwl-0.1/
+├─ ai_score.py          # human score / ai slop score
+├─ quickstart.py        # demo CLI
+├─ engine.py            # PipeOwl tokenizer + emb loader
+├─ pipeowl.safetensors  # embeddings + delta_field
+├─ tokenizer.json
+├─ ptt.npy              # style field
+├─ config.json
+├─ README.md
+├─ example.md
+└─ LICENSE
+```
+
+## LICENSE
+
 MIT