mlsanityData to 14 Oct 2026

Hybrid Adversarial Defence for Natural Language Understanding Tasks

arXiv 2606.04612 ↗First posted 3 Jun 2026cs.CLcited by 0

On the map: LLM safety and security → Jailbreaking

Abstract (arXiv)

Large Language Models (LLMs) are vulnerable both to hallucination and adversarial manipulation. Although these problems are closely related, existing defences typically address them separately. We investigate a hybrid defence framework that combines entropy-based models, designed to reduce hallucinations, with uncertainty-based models and geometric-based models, designed to reduce vulnerability. Under in-domain tests on Natural Language Understanding datasets (FEVER, HotpotQA, CSQA, SIQA) we find our hybrid model improves both clean-task performance (up to 43.34% increase in accuracy) and adversarial robustness (up to 64.92% improvement in accuracy and 62.27% reduction in attack success rate). For out-of-distribution datasets (AeroEngQA, CPIQA) we see similar adversarial robustness from our hybrid model (up to 57.14% improvement in accuracy). For prompt injection (SafeGuard) and jailbreak detection (AdvBench, DAN) datasets our hybrid model is also very strong (up to 51% reduction in attack success rate compared to state of the art baseline models). Overall, our results show that combining entropy, uncertainty and geometric features provides a more effective defence strategy than using any single feature alone for both in-domain and out-of-distribution tasks.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Jailbreaking, one of 12 directions in LLM safety and security. Builds on 33 of its 59 references that are in the map. First posted on arXiv 3 Jun 2026.

Builds on

33 of 59 references.

Closest ancestors · 4

Same direction, newest first

  1. 2025Bypassing LLM Guardrails: An Empirical Analysis of Evasion Attacks against Prompt Injection and Jailbreak Detection Systems15 Apr
  2. 2023A Wolf in Sheep's Clothing: Generalized Nested Jailbreak Prompts can Fool Large Language Models Easily14 Nov
  3. 2023Universal and Transferable Adversarial Attacks on Aligned Language Models27 Jul
  4. 2023Jailbroken: How Does LLM Safety Training Fail?5 Jul

Other references · 29

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  3. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
  4. 2014Explaining and Harnessing Adversarial Examplescited by 7,364
  5. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,754
  6. 2022Large Language Models Encode Clinical Knowledgecited by 1,830
  7. 2018CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledgecited by 1,822
  8. 2023Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injectioncited by 1,443
  9. 2018FEVER: a large-scale dataset for Fact Extraction and VERificationcited by 1,375
  10. 2019Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailmentcited by 671
  11. 2023The Internal State of an LLM Knows When It's Lyingcited by 659
  12. 2023Baseline Defenses for Adversarial Attacks Against Aligned Language Modelscited by 651
17 more
  1. 2023Safety-Tuned LLaMAs: Lessons From Improving the Safety of Large Language Models that Follow Instructionscited by 387
  2. 2018TextBugger: Generating Adversarial Text Against Real-world Applicationscited by 381
  3. 2019FreeLB: Enhanced Adversarial Training for Natural Language Understandingcited by 264
  4. 2023Do Large Language Models Know What They Don't Know?cited by 243
  5. 2023A Stitch in Time Saves Nine: Detecting and Mitigating Hallucinations of LLMs by Validating Low-Confidence Generationcited by 206
  6. 2023LLM Lies: Hallucinations are not Bugs, but Features as Adversarial Examplescited by 177
  7. 2022On the Origin of Hallucinations in Conversational Models: Is it the Datasets or the Models?cited by 138
  8. 2024Is LLM-as-a-Judge Robust? Investigating Universal Adversarial Attacks on Zero-shot LLM Assessmentcited by 126
  9. 2021Neural Path Hunter: Reducing Hallucination in Dialogue Systems via Path Groundingcited by 106
  10. 2025HalluLens: LLM Hallucination Benchmarkcited by 92
  11. 2021Towards Robustness Against Natural Language Word Substitutionscited by 69
  12. 2021Defending Pre-trained Language Models from Adversarial Word Substitutions Without Performance Sacrificecited by 30
  13. 2025Safe Pruning LoRA: Robust Distance-Guided Pruning for Safety Alignment in Adaptation of LLMscited by 10
  14. 2026Retrieval as Reasoning: Self-Evolving Agent-Native Retrieval via LLM-Wikicited by 9
  15. 2023DSRM: Boost Textual Adversarial Training with Distribution Shift Risk Minimizationcited by 7
  16. 2025SUGAR: Leveraging Contextual Confidence for Smarter Retrievalcited by 7
  17. 2025Adversarial Defence without Adversarial Defence: Enhancing Language Model Robustness via Instance-level Principal Component Removalcited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Attack and defense techniques in large language models: A survey and new perspectives
  2. 2024Breaking Down the Defenses: A Comparative Survey of Attacks on Large Language Models
  3. 2025AdversariaLLM: A Unified and Modular Toolbox for LLM Robustness Research
  4. 2023Survey of Vulnerabilities in Large Language Models Revealed by Adversarial Attacks
  5. 2025Tougher Text, Smarter Models: Raising the Bar for Adversarial Defence Benchmarks
  6. 2024Recent Advances in Attack and Defense Approaches of Large Language Models
  7. 2025Safeguarding Efficacy in Large Language Models: Evaluating Resistance to Human-Written and Algorithmic Adversarial Prompts
  8. 2024Jailbreak Attacks and Defenses Against Large Language Models: A Survey