Instructions to use ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward with Transformers:

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

pipe = pipeline("text-generation", model="ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward", trust_remote_code=True, device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.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(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward

SGLang

How to use ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward 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 "ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward" \
    --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": "ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward",
		"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 "ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward" \
        --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": "ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward with Docker Model Runner:
```
docker model run hf.co/ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward
```

PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward

A privacy-aligned variant of nvidia/NVIDIA-Nemotron-3-Nano-4B-BF16, fine-tuned with RL against an annotation-conditioned reward: a pairwise LLM judge that scores candidate responses conditioned on per-example privacy annotations.

Training

Method: online policy optimization (RL), KL-regularized to the base policy.
Algorithm: SAPO (tau_pos = 1.0, tau_neg = 1.05).
Reward: a pairwise LLM judge conditioned on the example's privacy annotations
Judge model: nvidia/NVIDIA-Nemotron-3-Nano-4B-BF16 (thinking enabled).

Hyperparameters


Base model	`nvidia/NVIDIA-Nemotron-3-Nano-4B-BF16`
Algorithm	SAPO
KL coefficient	0.002
Learning rate	1e-6 (constant w/ warmup, 10 step warmup)
Weight decay	0.01
Global prompt batch size	64
Precision	bf16
Num generations per prompt	4

Paper & citation

For full details on the method, data, and evaluation, please read our paper:

PrivacyAlign: Contextual Privacy Alignment for LLM Agents

@article{tamber2026privacyaligncontextualprivacyalignment,
      title={PrivacyAlign: Contextual Privacy Alignment for LLM Agents},
      author={Manveer Singh Tamber and Abhay Puri and Marc-Etienne Brunet and Perouz Taslakian and Jimmy Lin and Spandana Gella},
      year={2026},
      eprint={2606.21710},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2606.21710},
}

Downloads last month: 2

Model tree for ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward

Base model

nvidia/NVIDIA-Nemotron-Nano-12B-v2-Base

Finetuned

nvidia/NVIDIA-Nemotron-Nano-12B-v2

Finetuned

nvidia/NVIDIA-Nemotron-Nano-9B-v2

Finetuned

nvidia/NVIDIA-Nemotron-3-Nano-4B-BF16

Finetuned

(17)

this model

Paper for ServiceNow/PrivacyAlign-Nemotron-3-Nano-4B-Annotation-Conditioned-Reward

PrivacyAlign: Contextual Privacy Alignment for LLM Agents

Paper • 2606.21710 • Published Jun 19 • 4