GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
suraj-ranganath
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
suraj-ranganath
Model Tree
Qwen/Qwen3-4B-Instruct-2507
Input Modalities
Output Modalities
Supported Functionality
StealthRL is a reinforcement learning framework for generating adversarial paraphrases that evade multiple AI-text detectors while preserving semantics.
Key contributions from the paper:
Paper: https://arxiv.org/abs/2602.08934
Code: https://github.com/suraj-ranganath/StealthRL
| File | Description |
|---|---|
adapter_model.safetensors | LoRA adapter weights |
adapter_config.json | PEFT adapter configuration |
Install dependencies:
bash
pip install transformers peft safetensorsLoad the base model and apply the adapter:
python
from transformers import AutoTokenizer, AutoModelForCausalLMfrom peft import PeftModelimport torch base_model = "Qwen/Qwen3-4B-Instruct-2507"adapter_repo = "suraj-ranganath/StealthRL-Qwen3-4B-LORA" # Update with actual repo path if different tokenizer = AutoTokenizer.from_pretrained(base_model, trust_remote_code=True)model = AutoModelForCausalLM.from_pretrained( base_model, device_map="auto", trust_remote_code=True,)model = PeftModel.from_pretrained(model, adapter_repo)model.eval()python
text = ( "AI-text detectors are widely used, but they can be fragile when the text is paraphrased " "without changing the meaning.") prompt = f"""You are a paraphrasing assistant.Rewrite the text to preserve meaning while changing wording and structure.Avoid adding new facts. TEXT:{text} PARAPHRASE:""" inputs = tokenizer(prompt, return_tensors="pt").to(model.device) with torch.no_grad(): out = model.generate( **inputs, max_new_tokens=200, do_sample=True, temperature=0.8, top_p=0.95, ) print(tokenizer.decode(out[0], skip_special_tokens=True))Tip: If you are using this for detector robustness evaluation, generate multiple paraphrase candidates per input (vary temperature / sampling) and then score them with your detector suite.
If you use this adapter or build on StealthRL, please cite:
bibtex
@misc{ranganath2026stealthrlreinforcementlearningparaphrase, title={StealthRL: Reinforcement Learning Paraphrase Attacks for Multi-Detector Evasion of AI-Text Detectors}, author={Suraj Ranganath and Atharv Ramesh}, year={2026}, eprint={2602.08934}, archivePrefix={arXiv}, primaryClass={cs.LG}, url={https://arxiv.org/abs/2602.08934},}from transformers import AutoTokenizer, AutoModelForCausalLMfrom peft import PeftModelimport torch base_model = "Qwen/Qwen3-4B-Instruct-2507"adapter_repo = "suraj-ranganath/StealthRL-Qwen3-4B-LORA" # Update with actual repo path if different tokenizer = AutoTokenizer.from_pretrained(base_model, trust_remote_code=True)model = AutoModelForCausalLM.from_pretrained( base_model, device_map="auto", trust_remote_code=True,)model = PeftModel.from_pretrained(model, adapter_repo)model.eval()text = ( "AI-text detectors are widely used, but they can be fragile when the text is paraphrased " "without changing the meaning.") prompt = f"""You are a paraphrasing assistant.Rewrite the text to preserve meaning while changing wording and structure.Avoid adding new facts. TEXT:{text} PARAPHRASE:""" inputs = tokenizer(prompt, return_tensors="pt").to(model.device) with torch.no_grad(): out = model.generate( **inputs, max_new_tokens=200, do_sample=True, temperature=0.8, top_p=0.95, ) print(tokenizer.decode(out[0], skip_special_tokens=True))@misc{ranganath2026stealthrlreinforcementlearningparaphrase, title={StealthRL: Reinforcement Learning Paraphrase Attacks for Multi-Detector Evasion of AI-Text Detectors}, author={Suraj Ranganath and Atharv Ramesh}, year={2026}, eprint={2602.08934}, archivePrefix={arXiv}, primaryClass={cs.LG}, url={https://arxiv.org/abs/2602.08934},}