The Unseen Bias: How Norm Discrepancy in Pre-Norm MLLMs Leads to Visual Information Loss

Documenti analoghi

• GRAN-TED: Generating Robust, Aligned, and Nuanced Text Embedding for Diffusion Models
  di: Li, Bozhou, et al.
  Pubblicazione: (2025)
• Semantic Routing: Exploring Multi-Layer LLM Feature Weighting for Diffusion Transformers
  di: Li, Bozhou, et al.
  Pubblicazione: (2026)
• VidBridge-R1: Bridging QA and Captioning for RL-based Video Understanding Models with Intermediate Proxy Tasks
  di: Chen, Xinlong, et al.
  Pubblicazione: (2025)
• SEA: Supervised Embedding Alignment for Token-Level Visual-Textual Integration in MLLMs
  di: Yin, Yuanyang, et al.
  Pubblicazione: (2024)
• ID-Align: RoPE-Conscious Position Remapping for Dynamic High-Resolution Adaptation in Vision-Language Models
  di: Li, Bozhou, et al.
  Pubblicazione: (2025)