README.md

# CULTURE-MT: Beyond Literal Translation — Evaluating Cultural Effectiveness in Social Media UGC


[![Leaderboard](https://img.shields.io/badge/🏆-Leaderboard-yellow)](https://huggingface.co/spaces/Wulinjuan/CULTURE-MT)
[![License](https://img.shields.io/badge/License-CC%20BY%204.0-green)](https://creativecommons.org/licenses/by/4.0/)

CULTURE-MT is a benchmark for evaluating **CUL**tural **T**ransmission and **U**GC-specific emotion **RE**sonance in Chinese-to-English social media translation. It consists of 1,002 user-generated notes (UGC) spanning 14 content domains, presented at ICML 2026.

📄 **Paper:** [Beyond Literal Translation: Evaluating Cultural Effectiveness in Social Media UGC](https://arxiv.org/abs/2605.25626)

---

## 🌟 Why CULTURE-MT?

Standard machine translation metrics (BLEU, ChrF, COMET) fail to capture whether a translation truly resonates with target-language users. CULTURE-MT introduces **cultural effectiveness** as a new evaluation criterion, covering:

- **Expressive accuracy** — semantic fidelity, emotional tone, proper noun handling, and unit/measurement accuracy
- **Cultural adaptability** — culture-loaded term handling, overall cultural fluency, and addressing/politeness adaptation

---

## 📊 Dataset

| Split | Notes | Domains | Note Types |
|-------|-------|---------|------------|
| Benchmark | 1,002 | 14 | 4 (General, Express, Symbol, Hybrid) |

### Content Domains
Pets · Travel · Food · Crafts · Painting · Home Decoration · Outdoor · Sports · Fitness & Weight Loss · Technology & Gadgets · Cars · Games · Movies & TV · Celebrity News

### Note Types

| Type | Description |
|------|-------------|
| **General** | Informal UGC with few culture-loaded symbols or distinctive styles |
| **Express** | Strong rhetorical/expressive style (e.g., "planting grass", hyperbole, rhetorical questions) |
| **Symbol** | High density of internet cultural symbols (slang, memes, platform-specific jargon) |
| **Hybrid** | Both rich cultural symbols and distinctive expressive style |

---

## 📐 Evaluation

Translations are evaluated by **JUDGER**, a fine-tuned Qwen3-32B model trained on 30K expert- and LLM-annotated samples. It achieves 86.03% accuracy and Cohen's κ = 0.72 against human expert judgments.

### Scoring Rubric (0–3 scale)

| Score | Meaning |
|-------|---------|
| **0** | Severe meaning loss or distortion; target readers cannot grasp the original intent or emotion |
| **1** | Main idea barely understandable; critical cultural errors, poor adaptation |
| **2** | Main information conveyed accurately; reasonable emotional/contextual expression |
| **3** | Precise, natural, culturally fluent; fully conveys all information and emotion for English social media readers |

Scores 0–1 are treated as **culturally ineffective**; scores 2–3 as **culturally effective**.

---

## 🏆 Leaderboard

Submit your translations and get evaluated automatically by our trained **JUDGER** model at:
👉 https://huggingface.co/spaces/Wulinjuan/CULTURE-MT

> **📌 Note on Leaderboard vs. Paper Results**
> Scores reported in the ICML 2026 paper were produced with a single JUDGER inference pass (temperature = 0.6, top-p = 0.95). Since stochastic decoding introduces minor variation across runs, leaderboard scores are computed as the **average of four independent inference passes** under identical settings to ensure fairness and reproducibility. As a result, leaderboard scores may differ slightly from those reported in the paper.

### Top Results (as of 2026-05-27)

| Rank | Model | Ineff. ↓ | Eff. ↑ | 0 ↓ | 1 ↓ | 2 | 3 | Avg Score |
|------|-------|----------|--------|-----|-----|---|---|-----------|
| 1 | GPT-5 | 8.08% | 91.92% | 0.90% | 7.19% | 51.56% | 40.35% | 2.31 |
| 2 | Gemini 3 Pro | 8.95% | 91.05% | 0.20% | 8.75% | 51.97% | 39.09% | 2.30 |
| 3 | CULTURE-MT-baseline-32B | 12.08% | 87.92% | 0.73% | 11.35% | 56.56% | 31.37% | 2.19 |
| 4 | CULTURE-MT-baseline-8B | 14.34% | 85.66% | 1.27% | 13.07% | 57.42% | 28.24% | 2.13 |
| 5 | GLM4.6 | 15.94% | 84.06% | 3.12% | 12.81% | 57.39% | 26.68% | 2.08 |
| 6 | DeepSeek V3.2 | 17.73% | 82.27% | 2.59% | 15.14% | 56.82% | 25.45% | 2.05 |
| 7 | Qwen3-235B-A22B | 29.27% | 70.73% | 12.31% | 16.97% | 52.86% | 17.87% | 1.76 |
| 8 | Seed-X-PPO | 30.58% | 69.42% | 3.43% | 27.25% | 62.23% | 7.19% | 1.73 |

---

## 📤 How to Submit

Submissions are made directly through the leaderboard interface at:
👉 https://huggingface.co/spaces/Wulinjuan/CULTURE-MT

### Step 1 — Upload your translations

Upload a `submission.jsonl` file. Each line should be a JSON object with the `id` of the source note and your `translation`:

```json
{"id": "0001", "translation": "This place is amazing. I'll definitely come back!"}
{"id": "0002", "translation": "This is way too ridiculous."}
```

### Step 2 — Fill in model information

Complete the submission form with your model details (Model Name, Organization / Team, Base Model, Method, and a brief description). No additional files are required.

### Step 3 — Get your results

Aggregated scores will appear on the leaderboard automatically after evaluation. If you need detailed per-sample evaluation results, please contact us by email at **wulinjuan525@zju.edu.cn**.

---

## 📖 Citation

If you use CULTURE-MT in your research, please cite:

```bibtex
@misc{wu2026literaltranslationevaluatingcultural,
      title={Beyond Literal Translation: Evaluating Cultural Effectiveness in Social Media UGC}, 
      author={Linjuan Wu and Ruiqi Zhang and Xinze Lyu and Ye Guo and Daoxin Zhang and Zhe Xu and Yao Hu and Yixin Cao and Yongliang Shen and Weiming Lu},
      year={2026},
      eprint={2605.25626},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2605.25626}, 
}
```

---

## 📬 Contact

- Linjuan Wu: wulinjuan525@zju.edu.cn

This benchmark was developed with support from Zhejiang University, Fudan University, and Xiaohongshu Inc.