Cover Image

The Trigger in the Haystack: Extracting and Reconstructing LLM Backdoor Triggers

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Bullwinkel, Blake, Severi, Giorgio, Hines, Keegan, Minnich, Amanda, Kumar, Ram Shankar Siva, Zunger, Yonatan
Format:	Preprint
Published:	2026
Subjects:	Cryptography and Security Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

A Representation Engineering Perspective on the Effectiveness of Multi-Turn Jailbreaks
by: Bullwinkel, Blake, et al.
Published: (2025)

BadLLM-TG: A Backdoor Defender powered by LLM Trigger Generator
by: Zhang, Ruyi, et al.
Published: (2026)

PyRIT: A Framework for Security Risk Identification and Red Teaming in Generative AI System
by: Munoz, Gary D. Lopez, et al.
Published: (2024)

Revisiting Training-Inference Trigger Intensity in Backdoor Attacks
by: Lin, Chenhao, et al.
Published: (2025)

Invisible Textual Backdoor Attacks based on Dual-Trigger
by: Hou, Yang, et al.
Published: (2024)

ME: Trigger Element Combination Backdoor Attack on Copyright Infringement
by: Yang, Feiyu, et al.
Published: (2025)

Backdoor Attack with Invisible Triggers Based on Model Architecture Modification
by: Ma, Yuan, et al.
Published: (2024)

CL-Attack: Textual Backdoor Attacks via Cross-Lingual Triggers
by: Zheng, Jingyi, et al.
Published: (2024)

Evolutionary Trigger Detection and Lightweight Model Repair Based Backdoor Defense
by: Zhou, Qi, et al.
Published: (2024)

Pay Attention to the Triggers: Constructing Backdoors That Survive Distillation
by: De Muri, Giovanni, et al.
Published: (2025)

When Backdoors Go Beyond Triggers: Semantic Drift in Diffusion Models Under Encoder Attacks
by: Chen, Shenyang, et al.
Published: (2026)

Stealthy Dual-Trigger Backdoors: Attacking Prompt Tuning in LM-Empowered Graph Foundation Models
by: Xue, Xiaoyu, et al.
Published: (2025)

Defusing the Trigger: Plug-and-Play Defense for Backdoored LLMs via Tail-Risk Intrinsic Geometric Smoothing
by: Fan, Kaisheng, et al.
Published: (2026)

A Systematization of Security Vulnerabilities in Computer Use Agents
by: Jones, Daniel, et al.
Published: (2025)

Trusted Weights, Treacherous Optimizations? Optimization-Triggered Backdoor Attacks on LLMs
by: Wang, Yifei, et al.
Published: (2026)

Stealthy Backdoor Attacks against LLMs Based on Natural Style Triggers
by: Wei, Jiali, et al.
Published: (2026)

The Art of Deception: Robust Backdoor Attack using Dynamic Stacking of Triggers
by: Mengara, Orson
Published: (2024)

BadMoE: Backdooring Mixture-of-Experts LLMs via Optimizing Routing Triggers and Infecting Dormant Experts
by: Wang, Qingyue, et al.
Published: (2025)

Concealing Backdoor Model Updates in Federated Learning by Trigger-Optimized Data Poisoning
by: Zhang, Yujie, et al.
Published: (2024)

An LLM-Assisted Easy-to-Trigger Backdoor Attack on Code Completion Models: Injecting Disguised Vulnerabilities against Strong Detection
by: Yan, Shenao, et al.
Published: (2024)

Defending Against Indirect Prompt Injection Attacks With Spotlighting
by: Hines, Keegan, et al.
Published: (2024)

A Set of Generalized Components to Achieve Effective Poison-only Clean-label Backdoor Attacks with Collaborative Sample Selection and Triggers
by: Wu, Zhixiao, et al.
Published: (2025)

Paladin: Defending LLM-enabled Phishing Emails with a New Trigger-Tag Paradigm
by: Pang, Yan, et al.
Published: (2025)

From Shortcuts to Triggers: Backdoor Defense with Denoised PoE
by: Liu, Qin, et al.
Published: (2023)

Trapping Attacker in Dilemma: Examining Internal Correlations and External Influences of Trigger for Defending GNN Backdoors
by: Yang, Fan, et al.
Published: (2026)

Re-Triggering Safeguards within LLMs for Jailbreak Detection
by: Lin, Zheng, et al.
Published: (2026)

Discovering Universal Semantic Triggers for Text-to-Image Synthesis
by: Zhai, Shengfang, et al.
Published: (2024)

AutoBackdoor: Automating Backdoor Attacks via LLM Agents
by: Li, Yige, et al.
Published: (2025)

Universal and Context-Independent Triggers for Precise Control of LLM Outputs
by: Liang, Jiashuo, et al.
Published: (2024)

Securing Multi-turn Conversational Language Models From Distributed Backdoor Triggers
by: Tong, Terry, et al.
Published: (2024)

GRP-Obliteration: Unaligning LLMs With a Single Unlabeled Prompt
by: Russinovich, Mark, et al.
Published: (2026)

When Backdoors Speak: Understanding LLM Backdoor Attacks Through Model-Generated Explanations
by: Ge, Huaizhi, et al.
Published: (2024)

ShadowLogic: Backdoors in Any Whitebox LLM
by: Schulz, Kasimir, et al.
Published: (2025)

RESTRAIN: Reinforcement Learning-Based Secure Framework for Trigger-Action IoT Environment
by: Alam, Md Morshed, et al.
Published: (2025)

When Efficiency Backfires: Cascading LLMs Trigger Cascade Failure under Adversarial Attack
by: Sun, Zehan, et al.
Published: (2026)

Invisible to Humans, Triggered by Agents: Stealthy Jailbreak Attacks on Mobile Vision-Language Agents
by: Ding, Renhua, et al.
Published: (2025)

BadThink: Triggered Overthinking Attacks on Chain-of-Thought Reasoning in Large Language Models
by: Liu, Shuaitong, et al.
Published: (2025)

Exploring Backdoor Attack and Defense for LLM-empowered Recommendations
by: Ning, Liangbo, et al.
Published: (2025)

Collaborative Intelligence: Topic Modelling of Large Language Model use in Live Cybersecurity Operations
by: Lochner, Martin, et al.
Published: (2025)

Towards Sample-specific Backdoor Attack with Clean Labels via Attribute Trigger
by: Zhu, Mingyan, et al.
Published: (2023)