Pith. sign in

REVIEW 6 cited by

Test-Time Backdoor Attacks on Multimodal Large Language Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2402.08577 v1 pith:EVKK4TPG submitted 2024-02-13 cs.CL cs.CRcs.CVcs.LGcs.MM

classification cs.CLcs.CRcs.CVcs.LGcs.MM
keywords backdooranydoorattackseffectsharmfuluniversaladversarialdata
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Backdoor attacks are commonly executed by contaminating training data, such that a trigger can activate predetermined harmful effects during the test phase. In this work, we present AnyDoor, a test-time backdoor attack against multimodal large language models (MLLMs), which involves injecting the backdoor into the textual modality using adversarial test images (sharing the same universal perturbation), without requiring access to or modification of the training data. AnyDoor employs similar techniques used in universal adversarial attacks, but distinguishes itself by its ability to decouple the timing of setup and activation of harmful effects. In our experiments, we validate the effectiveness of AnyDoor against popular MLLMs such as LLaVA-1.5, MiniGPT-4, InstructBLIP, and BLIP-2, as well as provide comprehensive ablation studies. Notably, because the backdoor is injected by a universal perturbation, AnyDoor can dynamically change its backdoor trigger prompts/harmful effects, exposing a new challenge for defending against backdoor attacks. Our project page is available at https://sail-sg.github.io/AnyDoor/.

Discussion (0). Sign in to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Backdoor Attack on Vision Language Models with Stealthy Semantic Manipulation

    cs.CV 2025-06 conditional novelty 7.0 of 10

    BadSem shows that semantic mismatches between images and text can serve as stealthy backdoor triggers for VLMs, achieving near-perfect attack success with low poisoning rates.

  2. TokenSwap: Backdoor Attack on the Compositional Understanding of Large Vision-Language Models

    cs.CV 2025-09 conditional novelty 6.0 of 10

    TokenSwap poisons LVLMs so that triggered images produce captions with subject and object roles reversed, achieving high attack success while evading a perplexity-based detector.

  3. Semantic Router: On the Feasibility of Hijacking MLLMs via a Single Adversarial Perturbation

    cs.CV 2025-11 conditional novelty 5.0 of 10

    A single universal adversarial image perturbation can route different input semantics to different attacker-defined outputs in multimodal LLMs, with up to 66% success over five targets.

  4. Blockchain Network Analysis using Quantum Inspired Graph Neural Networks & Ensemble Models

    cs.LG 2025-08 unverdicted novelty 4.0 of 10

    The submission's abstract claims a quantum-inspired GNN with a CP-decomposition layer reaches 74.8% F2 on blockchain fraud detection, but the uploaded full text is an unrelated paper on VLM agent security.

  5. A Survey of Safety on Large Vision-Language Models: Attacks, Defenses and Evaluations

    cs.CR 2025-02 conditional novelty 4.0 of 10

    A survey of LVLM safety that adds a lifecycle taxonomy and new benchmark results showing Janus-Pro-7B has weaker safety than several open-source LVLMs.

  6. Backdoor Attacks and Defenses in Computer Vision Domain: A Survey

    cs.CR 2025-09 conditional novelty 3.0 of 10

    A literature review that classifies backdoor attacks and defenses in computer vision into a five-axis taxonomy and identifies supply-chain, hardware, and cross-task evaluation as open gaps.

Pith tools