Skip to content

Turn-Gate/turn-gate.github.io

Repository files navigation

TurnGate Research Website

arXiv Website GitHub code Cite Python

Official website for the paper TurnGate: Detecting Malicious Intent in Multi-turn Dialogue. TurnGate is a turn-level monitor that identifies the earliest turn where interaction becomes sufficient for harm, providing precise intervention while maintaining high utility for benign technical exploration.

Quick Start

python -m http.server 8080

# Open http://localhost:8080 in browser

Cite

If you find TurnGate useful in your research, please consider citing:

@misc{shen2026turnlateresponseawaredefense,
      title={One Turn Too Late: Response-Aware Defense Against Hidden Malicious Intent in Multi-Turn Dialogue}, 
      author={Xinjie Shen and Rongzhe Wei and Peizhi Niu and Haoyu Wang and Ruihan Wu and Eli Chien and Bo Li and Pin-Yu Chen and Pan Li},
      year={2026},
      eprint={2605.05630},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2605.05630}, 
}

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors