MET-D-Qwen3-4B-ms-only is a Malay-only moral reasoning model fine-tuned from Qwen3-4B. Given a moral dilemma, a character description, and a candidate action, it judges the action from that character's perspective and explains its judgment with an explicit chain-of-thought before answering. Moral dilemmas rarely have a single correct answer, which makes reasoning traces hard to verify. We address this by introducing a character perspective that yields a ground-truth answer, which is used for rejection-sampling the model's own reasoning traces, conditioned on a per-language, per-situation selection of theoretical grounds. Both the reasoning trace and the final answer are generated in Malay.

Model Details

Base model: Qwen/Qwen3-4B
Task: for a given (situation, character description, action) triple, answer two questions from the character's perspective:
1. Is the action acceptable? (Yes / No / Ambiguous)
2. Would (not) doing it be emotionally/mentally discomforting? (Yes / No)
Training data: self-generated reasoning traces, rejection-sampled against the ground truth per character perspective, in Malay only
Language coverage: this checkpoint is trained on Malay only. The multilingual model and other single-language variants are also available (see Model Variants below).
License: Apache 2.0 (inherited from the base model)

How to Get Started

python

from transformers import AutoTokenizer
from vllm import LLM, SamplingParams

model_id = "launch/MET-D-Qwen3-4B-ms-only"
tokenizer = AutoTokenizer.from_pretrained(model_id)
llm = LLM(model=model_id, max_model_len=16384)

situation = "..."            # the moral dilemma narrative
character = "..."            # the character description
action = "..."                # the candidate action to judge

prompt = f"""Saya akan memberikan anda dilema moral dan membentangkan anda dengan penerangan watak dan soalan.
Jawapan untuk soalan pertama hendaklah Ya, Tidak, atau Ambiguous, dan jawapan untuk soalan kedua hendaklah Ya atau Tidak.
Anda harus mematuhi format output dengan ketat, kerana ia akan dihuraikan secara automatik.
Pastikan anda mengeluarkan "Sebab:" dan "Jawapan:" untuk semua 2 soalan.

Dilema Moral:
{situation}

Perihalan Watak:
{character}

soalan:
1. Dari perspektif A, adakah tindakan berikut — sudah dijalankan, dianggap perlu diambil, boleh dianggap diambil, atau disoal dalam konteks situasi — dianggap boleh diterima? "{action}"
2. Jika jawapan sebelumnya ialah ya, adakah A mendapati melakukan tindakan itu secara mental atau emosi tidak selesa? Sebaliknya, jika jawapan sebelumnya adalah tidak, adakah A mendapati tidak melakukan tindakan itu secara mental atau emosi tidak selesa?

Jawapan anda:
1. Sebab: {{Sebab}} Jawapan: {{Ya/Tidak/Ambiguous}}
2. Sebab: {{Sebab}} Jawapan: {{Ya/Tidak}}
"""

chat_prompt = tokenizer.apply_chat_template(
    [{"role": "user", "content": prompt}],
    tokenize=False,
    add_generation_prompt=True,
)

sampling_params = SamplingParams(temperature=0.0, max_tokens=2048)
outputs = llm.generate(chat_prompt, sampling_params)
print(outputs[0].outputs[0].text)

Model Variants

This checkpoint is part of the MET collection, which includes the same task across base models and language subsets:

Repo	Base model	Language(s)
`launch/MET-D-Qwen3-4B`	Qwen3-4B	all 6 (mixed)
`launch/MET-D-Qwen3-4B-en-only`	Qwen3-4B	English only
`launch/MET-D-Qwen3-4B-es-only`	Qwen3-4B	Spanish only
`launch/MET-D-Qwen3-4B-hi-only`	Qwen3-4B	Hindi only
`launch/MET-D-Qwen3-4B-ko-only`	Qwen3-4B	Korean only
`launch/MET-D-Qwen3-4B-ms-only`	Qwen3-4B	Malay only
`launch/MET-D-Qwen3-4B-zh-only`	Qwen3-4B	Chinese only
`launch/MET-D-Qwen3-8B`	Qwen3-8B	all 6 (mixed)
`launch/MET-D-Qwen3-8B-en-only`	Qwen3-8B	English only
`launch/MET-D-Gemma3-4B`	Gemma-3-4B-it	all 6 (mixed)
`launch/MET-D-Gemma3-4B-en-only`	Gemma-3-4B-it	English only

Citation

If you use this, please cite:

bibtex

@article{lee2026met,
  title={MET: Theory-Grounded and Culture-Aware Multilingual Moral Reasoning},
  author={Lee, Ayoung and Kwon, Ryan and Zhang, Yunxiang and Liu, Yuxuan and Railton, Peter and Wang, Lu},
  journal={arXiv preprint arXiv:2607.11736},
  year={2026}
}