mlsanityData to 14 Oct 2026

STAR: SocioTechnical Approach to Red Teaming Language Models

arXiv 2406.11757 ↗First posted 17 Jun 2024cs.AI cs.CL cs.CYcited by 13

On the map: LLM safety and security → Prompt injection and agent security

Abstract (arXiv)

This research introduces STAR, a sociotechnical framework that improves on current best practices for red teaming safety of large language models. STAR makes two key contributions: it enhances steerability by generating parameterised instructions for human red teamers, leading to improved coverage of the risk surface. Parameterised instructions also provide more detailed insights into model failures at no increased cost. Second, STAR improves signal quality by matching demographics to assess harms for specific groups, resulting in more sensitive annotations. STAR further employs a novel step of arbitration to leverage diverse viewpoints and improve label reliability, treating disagreement not as noise but as a valuable contribution to signal quality.

Key facts. Cited by 13 papers in the map (data through 14 Oct 2026). Placed in Prompt injection and agent security, one of 12 directions in LLM safety and security. Builds on 17 of its 52 references that are in the map. First posted on arXiv 17 Jun 2024.

Builds on

17 of 52 references.

Closest ancestors · 6

Same direction, newest first

  1. 2024Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Prompts26 Feb
  2. 2024Red-Teaming for Generative AI: Silver Bullet or Security Theater?29 Jan
  3. 2023AART: AI-Assisted Red-Teaming with Diverse Data Generation for New LLM-powered Applications14 Nov
  4. 2023Red teaming ChatGPT via Jailbreaking: Bias, Robustness, Reliability and Toxicity30 Jan
  5. 2022Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned23 Aug
  6. 2022Red Teaming Language Models with Language Models7 Feb

Other references · 11

  1. 2023Universal and Transferable Adversarial Attacks on Aligned Language Modelscited by 3,104
  2. 2018UMAP: Uniform Manifold Approximation and Projection for Dimension Reductioncited by 2,898
  3. 2022LaMDA: Language Models for Dialog Applicationscited by 1,220
  4. 2023Low-Resource Languages Jailbreak GPT-4cited by 309
  5. 2024Black-Box Access is Insufficient for Rigorous AI Auditscited by 126
  6. 2023DICES Dataset: Diversity in Conversational AI Evaluation for Safetycited by 86
  7. 2022Is Your Toxicity My Toxicity? Exploring the Impact of Rater Identity on Toxicity Annotationcited by 69
  8. 2022Characteristics of Harmful Text: Towards Rigorous Benchmarking of Language Modelscited by 49
  9. 2024Aurora-M: Open Source Continual Pre-training for Multilingual Language and Codecited by 10
  10. 2023Intersectionality in Conversational AI Safety: How Bayesian Multilevel Models Help Understand Diverse Perceptions of Safetycited by 8
  11. 2023Adversarial Nibbler: A Data-Centric Challenge for Improving the Safety of Text-to-Image Modelscited by 7

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 3; 2025: 3; 2026: 7.

13 papers.

Sort:

By area:

  1. 2024Preference Tuning For Toxicity Mitigation Generalizes Across Languagescited by 31
  2. 2024Acceptable Use Policies for Foundation Modelscited by 12
  3. 2026Computational Hermeneutics: Evaluating generative AI as a cultural technologycited by 8
  4. 2024Policy Maps: Tools for Guiding the Unbounded Space of LLM Behaviorscited by 5
  5. 2025From Noise to Signal to Selbstzweck: Reframing Human Label Variation in the Era of Post-training in NLPcited by 3
  6. 2026Assessing Risks of Large Language Models in Mental Health Support: A Framework for Automated Clinical AI Red Teamingcited by 3
  7. 2026Formal Methods Meet LLMs: Auditing, Monitoring, and Intervention for Compliance of Advanced AI Systemscited by 3
  8. 2026Structuring the Space of Sociotechnical Alignmentcited by 2
  9. 2026Interactionless Inverse Reinforcement Learning: A Data-Centric Framework for Durable Alignmentcited by 1
  10. 2025"Just a strange pic": Evaluating 'safety' in GenAI Image safety annotation tasks from diverse annotators' perspectivescited by 0
  11. 2025InFerActive: Interactive Tree-Based Exploration of LLM Sampling for Safety Evaluationcited by 0
  12. 2026Long-term Measurements: Towards a Longitudinal Understanding of Human-AI Interactionscited by 0
1 more
  1. 2026A Translational Note on AI Safety Evaluationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023Explore, Establish, Exploit: Red Teaming Language Models from Scratch
  2. 2026Red Teaming LLMs as Socio-Technical Practice: From Exploration and Data Creation to Evaluation
  3. 2026CART: Closed-Loop Adaptive Red Teaming for Large Language Models
  4. 2025Building Safe GenAI Applications: An End-to-End Overview of Red Teaming for Large Language Models
  5. 2026RedVLA: Physical Red Teaming for Vision-Language-Action Models
  6. 2026STAR-Teaming: A Strategy-Response Multiplex Network Approach to Automated LLM Red Teaming
  7. 2024Embodied Red Teaming for Auditing Robotic Foundation Models
  8. 2025A Red Teaming Roadmap Towards System-Level Safety