Lossless Compression of Neural Network Components: Weights, Checkpoints, and K/V Caches in Low-Precision Formats

Documenti analoghi

• Understanding the Functional Roles of Modelling Components in Spiking Neural Networks
  di: Yin, Huifeng, et al.
  Pubblicazione: (2024)
• Set-based Neural Network Encoding Without Weight Tying
  di: Andreis, Bruno, et al.
  Pubblicazione: (2023)
• KV Cache Recycling to Expand Usable Context Capacity in Low Parameter LLMs
  di: Pandey, Prashant
  Pubblicazione: (2025)
• Gradient-Free Training of Spiking Neural Networks via Low-Rank Evolution Strategies
  di: Patankar, Dhruv, et al.
  Pubblicazione: (2026)
• LANCE: Low Rank Activation Compression for Efficient On-Device Continual Learning
  di: Apolinario, Marco Paul E., et al.
  Pubblicazione: (2025)