Beyond What Meets the Eye: Unveiling Situational Illusions for Multimodal Large Language Models

arXiv:2608.22232 · cs.AI, cs.CL, cs.CV, cs.MM · Submitted 2026-08-23 · Read on arXiv

cs.AI, cs.CL, cs.CV, cs.MM

Submitted: 2026-08-23

Updated: 2026-08-25

Comments: 9 pages, 5 figures

Code: https://github.com/yangzm1105/MSIBench

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Accuracy — 13× in this paper · explained in 1 episode
• Benchmark — 11× in this paper · explained in 2 episodes
• Grounding — 10× in this paper · explained in 3 episodes
• Failure Modes — 8× in this paper · explained in 3 episodes
• Benchmarking — 5× in this paper · explained in 3 episodes
• Fine-Tuning — 5× in this paper · explained in 5 episodes
• Contextual Understanding — 3× in this paper · explained in 2 episodes
• Core — 3× in this paper · explained in 2 episodes
• LLM-as-a-Judge — 3× in this paper · explained in 3 episodes
• Robustness — 3× in this paper · explained in 5 episodes
• Multimodal Reasoning — 2× in this paper · explained in 5 episodes
• Supervised Fine-Tuning (SFT) — 2× in this paper · explained in 8 episodes
• Task Planning — 2× in this paper · explained in 2 episodes
• Active Exploration — 1× in this paper · explained in 2 episodes
• Hallucination — 1× in this paper · explained in 3 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Verification — 1× in this paper · explained in 2 episodes

Sources

• Qwen3-VL Technical Report
• CHARTOM: A Visual Theory-of-Mind Benchmark for LLMs on Misleading Charts
• MVI-Bench: A Comprehensive Benchmark for Evaluating Robustness to Misleading Visual Inputs in LVLMs
• HomeSafeBench: Benchmarking Embodied Vision-Language Models in Free-Exploration Home Safety Inspection
• Gemma 3 Technical Report
• GLM-4.5V and GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning
• Seeing Is Believing? A Benchmark for Multimodal Large Language Models on Visual Illusions and Anomalies
• A Framework for Benchmarking and Aligning Task-Planning Safety in LLM-Based Embodied Agents
• Kimi-VL Technical Report
• Kimi K2.5: Visual Agentic Intelligence
• MMR-Life: Piecing Together Real-life Scenes for Multimodal Multi-image Reasoning
• AGENTSAFE: Benchmarking the Safety of Embodied Agents on Hazardous Instructions
• IllusionVQA: A Challenging Optical Illusion Dataset for Vision Language Models
• InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency
• SafeAgentBench: A Benchmark for Safe Task Planning of Embodied LLM Agents

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems
• VSAL: A Vision Solver with Adaptive Layouts for Graph Property Detection