STILL: Selecting Tokens for Intra-Layer Hybrid Attention to Linearize LLMs

Documenti analoghi

• MirrorLA: Reflecting Feature Map for Vision Linear Attention
  di: Meng, Weikang, et al.
  Pubblicazione: (2026)
• Norm$\times$Direction: Restoring the Missing Query Norm in Vision Linear Attention
  di: Meng, Weikang, et al.
  Pubblicazione: (2025)
• PolaFormer: Polarity-aware Linear Attention for Vision Transformers
  di: Meng, Weikang, et al.
  Pubblicazione: (2025)
• Neural Attention Search Linear: Towards Adaptive Token-Level Hybrid Attention Models
  di: Deng, Difan, et al.
  Pubblicazione: (2026)
• Alleviating Forgetfulness of Linear Attention by Hybrid Sparse Attention and Contextualized Learnable Token Eviction
  di: He, Mutian, et al.
  Pubblicazione: (2025)