| title | Stable-GFlowNet: Toward Diverse and Robust LLM Red-Teaming via Contrastive Trajectory Balance |
|---|---|
| authors | Minchan Kwon, Sunghyun Baek, Minseo Kim, Jaemyung Yu, Dongyoon Han, Junmo Kim |
| date | 2026-05-01 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2605.00553 |
| upvotes | 14 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf |
| abstract | Large Language Model (LLM) Red-Teaming, which proactively identifies vulnerabilities of LLMs, is an essential process for ensuring safety. Finding effective and diverse attacks in red-teaming is important, but achieving both is challenging. Generative Flow Networks (GFNs) that perform distribution matching are a promising methods, but they are notorious for training instability and mode collapse. In particular, unstable rewards in red-teaming accelerate mode collapse. We propose Stable-GFN (S-GFN), which eliminates partition function Z estimation in GFN and reduces training instability. S-GFN avoids Z-estimation through pairwise comparisons and employs a robust masking methodology against noisy rewards. Additionally, we propose a fluency stabilizer to prevent the model from getting stuck in local optima that produce gibberish. S-GFN provides more stable training while maintaining the optimal policy of GFN. We demonstrate the overwhelming attack performance and diversity of S-GFN across various settings. |
cite
@misc{kwon2026stable,
title={Stable-GFlowNet: Toward Diverse and Robust LLM Red-Teaming via Contrastive Trajectory Balance},
author={Minchan Kwon and Sunghyun Baek and Minseo Kim and Jaemyung Yu and Dongyoon Han and Junmo Kim},
year={2026},
eprint={2605.00553},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2605.00553}
}