• Home
  • Search
  • Ctta: a novel chain-of-thought transfer adversarial attacks framework for large language models
  • Cite Icon3
  • https://doi.org/10.1186/s42400-024-00338-1Copy DOI Icon

Ctta: a novel chain-of-thought transfer adversarial attacks framework for large language models

Show More
  • Abstract
  • Highlights & Summary
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

Recent studies have indicated that large language models (LLMs) remain susceptible to adversarial attacks, despite enhanced robustness through the chain-of-thought (CoT) capability. However, this capability also introduces the potential for more covert and effective adversarial attack methods. This paper proposes a CoT Transfer Adversarial attack framework (CTTA) for general LLMs. Initially, we utilize a pre-trained model based on the transformer architecture and fine-tune it on various tasks to serve as a surrogate model. Subsequently, different levels of adversarial attack algorithms are utilized, and the generated adversarial samples are used as transfer samples. A thought chain-based adversarial transfer attack framework is constructed using transfer samples and thought chain techniques. Finally, various indicators are utilized to assess the performance of the general LLMs in response to this attack. The results demonstrate that the attack framework surpasses current state-of-the-art research. Numerous experiments on LLMs with varying performance and parameter sizes have validated the effectiveness, stability, and generalizability of this attack. The model’s error response and the superiority of this attack are thoroughly examined using attention by gradient technology, confirming the security threats posed by LLMs when leveraging CoT capability. This has significant implications for enhancing the security and robustness of LLMs.

Similar Papers
  • Research Article

Exploring Vulnerabilities and Threats in Large Language Models: Safeguarding Against Exploitation and Misuse

  • Aug 10, 2024
  • INTERANTIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT
  • Mr Aarush Varma +1
  • Conference Article

CrossLM: A Data-Free Collaborative Fine-Tuning Framework for Large and Small Language Models

  • Jun 23, 2025
  • Yongheng Deng +5
  • Research Article
  • Citations114

Multi-model assurance analysis showing large language models are highly vulnerable to adversarial hallucination attacks during clinical decision support

  • Aug 02, 2025
  • Communications Medicine
  • Mahmud Omar +10
  • Research Article
  • Citations17

Transferable adversarial distribution learning: Query-efficient adversarial attack against large language models

  • Sep 18, 2023
  • Computers & Security
  • Huoyuan Dong +4
  • Research Article
  • Citations14

ROFED-LLM: Robust Federated Learning for Large Language Models in Adversarial Wireless Environments

  • Jan 01, 2026
  • IEEE Transactions on Network Science and Engineering
  • Haoyu Wang +6
  • Research Article
  • Citations1

Beyond vulnerabilities: A comprehensive survey of adversarial attacks across domains

  • Aug 01, 2026
  • Computer Science Review
  • Dimitrios Christos Asimopoulos +3
  • PDF
  • Research Article
  • Citations1

Research on adversarial attack and defense of large language models

  • Nov 08, 2024
  • Applied and Computational Engineering
  • Jidong Yang +3
  • Research Article

Can adversarial attacks by large language models be attributed?

  • Feb 23, 2026
  • PLOS Complex Systems
  • Manuel Cebrian +2
  • Conference Article
  • Citations29

Is LLM-as-a-Judge Robust? Investigating Universal Adversarial Attacks on Zero-shot LLM Assessment

  • Jan 01, 2024
  • Vyas Raina +2
  • Research Article

Athreat Modeling Framework for LLM System Integration Leveraging NLP and Machine Learning

  • Apr 29, 2025
  • Communications on Applied Nonlinear Analysis
  • Basavraj Gadade, G A Patil
  • Research Article
  • Citations7

Mind Mapping Prompt Injection: Visual Prompt Injection Attacks in Modern Large Language Models

  • May 08, 2025
  • Electronics
  • Seyong Lee +2
  • Research Article
  • Citations1

Drug or Pokémon? An analysis of the ability of large language models to discern fabricated medications

  • Jan 13, 2026
  • medRxiv
  • Kelli Henry +14
  • Research Article
  • Citations4

Adversarial attack method against image classification based on haze perturbation

  • Feb 01, 2023
  • SCIENTIA SINICA Informationis
  • 伟 冯 +3
  • Research Article
  • Citations24

Red Teaming Language Model Detectors with Language Models

  • Feb 23, 2024
  • Transactions of the Association for Computational Linguistics
  • Zhouxing Shi +5
  • Research Article
  • Citations256

Adversarial Machine Learning Attacks and Defense Methods in the Cyber Security Domain

  • May 25, 2021
  • ACM Computing Surveys
  • Ishai Rosenberg +3
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.