Instructions to use SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct")
model = AutoModelForImageTextToText.from_pretrained("SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct

SGLang

How to use SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct with Docker Model Runner:
```
docker model run hf.co/SCAI-JHU/MindZero-gw-asst-Qwen3-VL-8B-Instruct
```

MindZero-gw-asst-Qwen3-VL-8B-Instruct

A MindZero checkpoint trained from Qwen/Qwen3-VL-8B-Instruct with self-supervised reinforcement learning for online proactive assistance in GridWorld environments.

TL;DR

MindZero trains (M)LLMs to perform efficient and robust online mental reasoning without any mental-state annotations. During training, the model is rewarded for generating mental-state hypotheses that maximize the likelihood of observed actions, as estimated by a planner — analogous to model-based ToM reasoning. After training, MindZero internalizes this reasoning into fast single-pass inference.

Evaluation

Base model	Checkpoint	Speedup on GridWorld Proactive Assistance
Qwen/Qwen3-VL-4B-Instruct	MindZero-gw-asst-Qwen3-VL-4B-Instruct	23.0
Qwen/Qwen3-VL-8B-Instruct	MindZero-gw-asst-Qwen3-VL-8B-Instruct	24.5

Citation

@inproceedings{zhang2026mindzero,
  title     = {MindZero: Learning Online Mental Reasoning With Zero Annotations},
  author    = {Shunchi Zhang and Jin Lu and Chuanyang Jin and Yichao Zhou and Zhining Zhang and Tianmin Shu},
  booktitle = {Proceedings of the 43rd International Conference on Machine Learning (ICML)},
  year      = {2026}
}