Beyond IID weights: sparse and low-rank deep Neural Networks are also Gaussian Processes

Documenti analoghi

• Mind the Gap: a Spectral Analysis of Rank Collapse and Signal Propagation in Attention Layers
  di: Saada, Thiziri Nait, et al.
  Pubblicazione: (2024)
• A simple proof of almost sure convergence for the largest singular value of a product of Gaussian matrices
  di: Saada, Thiziri Nait, et al.
  Pubblicazione: (2024)
• The Data-Quality Illusion: Rethinking Classifier-Based Quality Filtering for LLM Pretraining
  di: Saada, Thiziri Nait, et al.
  Pubblicazione: (2025)
• GraphEval: A Knowledge-Graph Based LLM Hallucination Evaluation Framework
  di: Sansford, Hannah, et al.
  Pubblicazione: (2024)
• Theory of Minimal Weight Perturbations in Deep Networks and its Applications for Low-Rank Activated Backdoor Attacks
  di: Evans, Bethan, et al.
  Pubblicazione: (2026)