Globally optimized SVD compression of LLMs via Fermi-function-based rank selection and gauge fixing

Documenti analoghi

• Only relative ranks matter in weight-clustered large language models
  di: Aizpurua, Borja, et al.
  Pubblicazione: (2026)
• Tensor network compressibility of convolutional models
  di: Singh, Sukhbinder, et al.
  Pubblicazione: (2024)
• Tensorization is a powerful but underexplored tool for compression and interpretability of neural networks
  di: Hamreras, Safa, et al.
  Pubblicazione: (2025)
• Fast Tensorization of Neural Networks via Slice-wise Feature Distillation
  di: Hamreras, Safa, et al.
  Pubblicazione: (2026)
• Block removal for large language models through constrained binary optimization
  di: Jansen, David, et al.
  Pubblicazione: (2026)