Variance Is Not Importance: Structural Analysis of Transformer Compressibility Across Model Scales

↘ Related papers

• Working Memory Constraints Scaffold Learning in Transformers under Data Scarcity
  Pranava Madhyastha48% match
• Explicit Dropout: Deterministic Regularization for Transformer Architectures
  Vidhi Agrawal45% match
• GSQ: Highly-Accurate Low-Precision Scalar Quantization for LLMs via Gumbel-Softmax Sampling
  Alireza Dadgarnia37% match
• AAC: Admissible-by-Architecture Differentiable Landmark Compression for ALT
  An T. Le36% match
• Convergent Evolution: How Different Language Models Learn Similar Number Representations
  Deqing Fu35% match