@inproceedings{ziakas-etal-2026-red,
    title = "Red-Bandit: Test-Time Adaptation for {LLM} Red-Teaming via Bandit-Guided {L}o{RA} Experts",
    author = "Ziakas, Christos  and
      Loo, Nicholas  and
      Jain, Nishita  and
      Russo, Alessandra",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-long.2156/",
    pages = "46462--46478",
    ISBN = "979-8-89176-390-6",
    abstract = "Automated red-teaming has emerged as a scalable approach for auditing Large Language Models (LLMs) prior to deployment, yet existing approaches lack mechanisms to efficiently adapt to model-specific vulnerabilities at inference. We introduce Red-Bandit, a red-teaming framework that adapts online to identify and exploit model failure modes under distinct attack styles (e.g., manipulation, slang). Red-Bandit post-trains a set of parameter-efficient LoRA experts, each specialized for a particular attack style, using reinforcement learning that rewards the generation of unsafe prompts via a rule-based safety model. At inference, a multi-armed bandit policy dynamically selects among these attack-style experts based on the target model{'}s response safety, balancing exploration and exploitation. Red-Bandit outperforms state-of-the-art methods on AdvBench and HarmBench, achieving higher attack success rates under sufficient exploration budgets (ASR@10), while generating more human-readable adversarial prompts (lower perplexity). In addition, Red-Bandit{'}s bandit policy serves as a diagnostic tool for identifying model-specific vulnerabilities by indicating which attack styles most effectively elicit harmful behaviors."
}