mlsanityData to 14 Oct 2026

MANATEE: Inference-Time Lightweight Diffusion Based Safety Defense for LLMs

arXiv 2602.18782 ↗First posted 21 Feb 2026cs.CR cs.AI cs.CLcited by 2

On the map: LLM safety and security → Jailbreaking

Abstract (arXiv)

Defending LLMs against adversarial jailbreak attacks remains an open challenge. Existing defenses rely on binary classifiers that fail when adversarial input falls outside the learned decision boundary, and repeated fine-tuning is computationally expensive while potentially degrading model capabilities. We propose MANATEE, an inference-time defense that uses density estimation over a benign representation manifold. MANATEE learns the score function of benign hidden states and uses diffusion to project anomalous representations toward safe regions--requiring no harmful training data and no architectural modifications. Experiments across Mistral-7B-Instruct, Llama-3.1-8B-Instruct, and Gemma-2-9B-it demonstrate that MANATEE reduce Attack Success Rate by up to 100% on certain datasets, while preserving model utility on benign inputs.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in Jailbreaking, one of 12 directions in LLM safety and security. Builds on 19 of its 25 references that are in the map. First posted on arXiv 21 Feb 2026.

Builds on

19 of 25 references.

Closest ancestors · 2

Same direction, newest first

  1. 2023Bypassing the Safety Training of Open-Source LLMs with Priming Attacks19 Dec
  2. 2023Universal and Transferable Adversarial Attacks on Aligned Language Models27 Jul

Other references · 17

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2024The Llama 3 Herd of Modelscited by 14,715
  3. 2017On Calibration of Modern Neural Networkscited by 4,960
  4. 2023Mistral 7Bcited by 3,164
  5. 2022Locating and Editing Factual Associations in GPTcited by 2,691
  6. 2016A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networkscited by 2,385
  7. 2019Fine-Tuning Language Models from Human Preferencescited by 2,150
  8. 2024Gemma 2: Improving Open Language Models at a Practical Sizecited by 1,998
  9. 2018A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attackscited by 1,485
  10. 2020Energy-based Out-of-distribution Detectioncited by 1,299
  11. 2023Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!cited by 1,154
  12. 2017Selective Classification for Deep Neural Networkscited by 852
5 more
  1. 2024Training Large Language Models to Reason in a Continuous Latent Spacecited by 797
  2. 2024Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Trainingcited by 529
  3. 2024Improving Instruction-Following in Language Models through Activation Steeringcited by 160
  4. 2025The Rogue Scalpel: Activation Steering Compromises LLM Safetycited by 21
  5. 2025Mechanistic Anomaly Detection for "Quirky" Language Modelscited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026Density Ridge Selective Prediction for LLM and VLM Hallucination Detection under Calibration Label Scarcitycited by 1
  2. 2026Active Adaptation, Not Static Defense: Temporal Dynamics of Preventative Steering in Adversarial Fine-Tuningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026SafeHarbor: Defining Precise Decision Boundaries via Hierarchical Memory-Augmented Guardrail for LLM Agent Safety
  2. 2026Model-Agnostic Lifelong LLM Safety via Externalized Attack-Defense Co-Evolution
  3. 2026Safeguarding LLMs via Model-Agnostic Latent Safety Signals from Dark Knowledge
  4. 2026BarrierSteer: LLM Safety via Learning Barrier Steering
  5. 2026Inference-Layer Security: Defending Against Adversarial Inference and Infrastructure Abuse
  6. 2024LIAR: Leveraging Inference Time Alignment (Best-of-N) to Jailbreak LLMs in Seconds
  7. 2026Ellipsoid Control: A White-list Jailbreak Defense via Benign Latent Modeling
  8. 2025Diffusion LLMs are Natural Adversaries for any LLM