MMBench

MMBench is a multilingual benchmark derived from widely used English datasets, designed to evaluate the performance of vision-language models across five European languages.

Papers using MMBench (32)

STEP3-VL-10B Technical Report2026MotionVLA: Vision-Language-Action Model for Humanoid Motion2026Id-align: Rope-conscious Position Remapping For Dynamic High-resolution Adaptation In Vision-language Models2025Multilingual Training and Evaluation Resources for Vision-Language Models2026VLM-RobustBench: A Comprehensive Benchmark for Robustness of Vision-Language Models2026Cheers: Decoupling Patch Details from Semantic Representations Enables Unified Multimodal Comprehension and Generation2026Perceptio: Perception Enhanced Vision Language Models via Spatial Token Generation2026ACPO: Counteracting Likelihood Displacement in Vision-Language Alignment with Asymmetric Constraints2026Annotation-Efficient Vision-Language Model Adaptation to the Polish Language Using the LLaVA Framework2026How to Take a Memorable Picture? Empowering Users with Actionable Feedback2026Text-Guided Layer Fusion Mitigates Hallucination in Multimodal LLMs2026Scaling Vision Language Models for Pharmaceutical Long Form Video Reasoning on Industrial GenAI Platform2026Gated Relational Alignment via Confidence-based Distillation for Efficient VLMs2026Stitch and Tell: A Structured Multimodal Data Augmentation Method for Spatial Understanding2025HybridToken-VLM: Hybrid Token Compression for Vision-Language Models2025AttAnchor: Guiding Cross-Modal Token Alignment in VLMs with Attention Anchors2025The Telephone Game: Evaluating Semantic Drift in Unified Models2025Self-Consistency as a Free Lunch: Reducing Hallucinations in Vision-Language Models via Self-Reflection2025Language-specific Layer Matters: Efficient Multilingual Enhancement For Large Vision-language Models2025Gam-agent: Game-theoretic And Uncertainty-aware Collaboration For Complex Visual Reasoning2025PostAlign: Multimodal Grounding as a Corrective Lens for MLLMs2025From Evaluation to Defense: Advancing Safety in Video Large Language Models2025MMTBENCH: A Unified Benchmark for Complex Multimodal Table Reasoning2025MOVE: A Mixture-of-Vision-Encoders Approach for Domain-Focused Vision-Language Processing2025MMBench: Is Your Multi-modal Model an All-around Player?2023 · 33 citesInternLM-XComposer: A Vision-Language Large Model for Advanced Text-image Comprehension and Composition2023 · 31 citesMMICL: Empowering Vision-language Model with Multi-Modal In-Context Learning2023 · 18 citesEE-MLLM: A Data-Efficient and Compute-Efficient Multimodal Large Language Model2024 · 1 citesRetrieval Meets Reasoning: Even High-school Textbook Knowledge Benefits Multimodal Reasoning2024Dynamic Multimodal Evaluation with Flexible Complexity by Vision-Language Bootstrapping2024Enhancing Instruction-Following Capability of Visual-Language Models by Reducing Image Redundancy2024Enhancing Vision-Language Model Reliability with Uncertainty-Guided Dropout Decoding2024