Both Text and Images Leaked! A Systematic Analysis of Data Contamination in Multimodal LLM

Documenti analoghi

• LongLLaVA: Scaling Multi-modal LLMs to 1000 Images Efficiently via a Hybrid Architecture
  di: Wang, Xidong, et al.
  Pubblicazione: (2024)
• Less is More: A Simple yet Effective Token Reduction Method for Efficient Multi-modal LLMs
  di: Song, Dingjie, et al.
  Pubblicazione: (2024)
• Can Multimodal LLMs See Materials Clearly? A Multimodal Benchmark on Materials Characterization
  di: Lai, Zhengzhao, et al.
  Pubblicazione: (2025)
• Multimodal LLM-Guided Semantic Correction in Text-to-Image Diffusion
  di: Lv, Zheqi, et al.
  Pubblicazione: (2025)
• Harmfully Manipulated Images Matter in Multimodal Misinformation Detection
  di: Wang, Bing, et al.
  Pubblicazione: (2024)