mlsanityData to 7 Oct 2026

On the Vulnerability of LLM/VLM-Controlled Robotics

arXiv 2402.10340 ↗First posted 15 Feb 2024cs.RO cs.AIcited by 15

On the map: Multimodal LLMs / vision-language models → Embodied and 3D vision-language models

Abstract (arXiv)

In this work, we highlight vulnerabilities in robotic systems integrating large language models (LLMs) and vision-language models (VLMs) due to input modality sensitivities. While LLM/VLM-controlled robots show impressive performance across various tasks, their reliability under slight input variations remains underexplored yet critical. These models are highly sensitive to instruction or perceptual input changes, which can trigger misalignment issues, leading to execution failures with severe real-world consequences. To study this issue, we analyze the misalignment-induced vulnerabilities within LLM/VLM-controlled robotic systems and present a mathematical formulation for failure modes arising from variations in input modalities. We propose empirical perturbation strategies to expose these vulnerabilities and validate their effectiveness through experiments on multiple robot manipulation tasks. Our results show that simple input perturbations reduce task execution success rates by 22.2% and 14.6% in two representative LLM/VLM-controlled robotic systems. These findings underscore the importance of input modality robustness and motivate further research to ensure the safe and reliable deployment of advanced LLM/VLM-controlled robotic systems.

Key facts. Cited by 15 papers in the map (data through 7 Oct 2026). Placed in Embodied and 3D vision-language models, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 29 of its 59 references that are in the map. First posted on arXiv 15 Feb 2024.

Builds on

29 of 59 references.

Closest ancestors · 2

Same direction, newest first

  1. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control28 Jul
  2. 2022Visual Language Maps for Robot Navigation11 Oct

Other references · 27

  1. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  2. 2023GPT-4 Technical Reportcited by 18,132
  3. 2023Visual Instruction Tuningcited by 7,792
  4. 2013Intriguing properties of neural networkscited by 5,114
  5. 2022Flamingo: a Visual Language Model for Few-Shot Learningcited by 4,337
  6. 2023Universal and Transferable Adversarial Attacks on Aligned Language Modelscited by 2,914
  7. 2022Do As I Can, Not As I Say: Grounding Language in Robotic Affordancescited by 2,302
  8. 2023Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injectioncited by 1,339
  9. 2022Code as Policies: Language Model Programs for Embodied Controlcited by 1,172
  10. 2023The Dawn of LMMs: Preliminary Explorations with GPT-4V(ision)cited by 677
  11. 2023Baseline Defenses for Adversarial Attacks Against Aligned Language Modelscited by 608
  12. 2023HallusionBench: An Advanced Diagnostic Suite for Entangled Language Hallucination and Visual Illusion in Large Vision-Language Modelscited by 572
15 more
  1. 2022Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulationcited by 545
  2. 2022VIMA: General Robot Manipulation with Multimodal Promptscited by 403
  3. 2023Language to Rewards for Robotic Skill Synthesiscited by 307
  4. 2023Automatically Auditing Large Language Models via Discrete Optimizationcited by 187
  5. 2020Improving Vision-and-Language Navigation with Image-Text Pairs from the Webcited by 177
  6. 2023A Survey of Large Language Models for Healthcare: from Data, Technology, and Applications to Accountability and Ethicscited by 162
  7. 2022The Unsurprising Effectiveness of Pre-Trained Vision Models for Controlcited by 155
  8. 2023MMC: Advancing Multimodal Chart Understanding with Large-scale Instruction Tuningcited by 149
  9. 2023Instruct2Act: Mapping Multi-modality Instructions to Robotic Actions with Large Language Modelcited by 115
  10. 2023Can an Embodied Agent Find Your "Cat-shaped Mug"? LLM-Guided Exploration for Zero-Shot Object Navigationcited by 74
  11. 2023MUTEX: Learning Unified Policies from Multimodal Task Specificationscited by 70
  12. 2022LATTE: LAnguage Trajectory TransformErcited by 40
  13. 2023COVID-VTS: Fact Extraction and Verification on Short Video Platformscited by 15
  14. 2023MTG: Mapless Trajectory Generator with Traversability Coverage for Outdoor Navigationcited by 10
  15. 2023Safety Alignment in NLP Tasks: Weakly Aligned Summarization as an In-Context Attackcited by 10

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 3; 2025: 4; 2026: 8.

15 papers.

Sort:

By area:

  1. 2024A Survey of Attacks on Large Vision-Language Models: Resources, Advances, and Future Trendscited by 82
  2. 2024AutoHallusion: Automatic Generation of Hallucination Benchmarks for Vision-Language Modelscited by 30
  3. 2024LOC-ZSON: Language-driven Object-Centric Zero-Shot Object Retrieval and Navigationcited by 8
  4. 2026SABER: A Stealthy Agentic Black-Box Attack Framework for Vision-Language-Action Modelscited by 2
  5. 2026Learning Robust Execution in Robotic Manipulation with Agentic Reinforcement Learningcited by 2
  6. 2025Unfettered Forceful Skill Acquisition with Physical Reasoning and Coordinate Frame Labelingcited by 1
  7. 2025AERMANI-VLM: Structured Prompting and Reasoning for Aerial Manipulation with Vision Language Modelscited by 1
  8. 2026Benchmarking the Safety of Large Language Models for Robotic Health Attendant Controlcited by 1
  9. 2025On the Dual-Use Dilemma in Physical Reasoning and Forcecited by 0
  10. 2025DRIP: Dynamic patch Reduction via Interpretable Poolingcited by 0
  11. 2026Revealing Interpretable Failure Modes of VLMscited by 0
  12. 2026Probing Collision Grounding in Vision-Language Models for Safe Human-Robot Collaborationcited by 0
3 more
  1. 2026Security of Foundation-Model-Powered Embodied Agents: Attack Surfaces, Attacks, Defenses, and Evaluationcited by 0
  2. 2026Generative Interpretability via Scalable Neuro-Symbolic Modelscited by 0
  3. 2026Security and Privacy in Large-Model-Driven Embodied Agents: Attacks, Defenses, and Future Directionscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025LIBERO-Plus: In-depth Robustness Analysis of Vision-Language-Action Models
  2. 2026Is Success All You Need? Investigating the Impact of Input Perturbations on VLA Behaviour in Tabletop Manipulation Tasks
  3. 2025RoboInspector: Unveiling the Unreliability of Policy Code for LLM-enabled Robotic Manipulation
  4. 2025Adversarial Attacks on Robotic Vision Language Action Models
  5. 2025Large VLM-based Vision-Language-Action Models for Robotic Manipulation: A Survey
  6. 2025REALM: A Real-to-Sim Validated Benchmark for Generalization in Robotic Manipulation
  7. 2024What Matters in Building Vision-Language-Action Models for Generalist Robots
  8. 2025Eva-VLA: Evaluating Vision-Language-Action Models' Robustness Under Real-World Physical Variations