A Mesoscopic View of Transformer Weights Through Row and Column Scale Fields
arXiv:2609.35852v1 Announce Type: cross Abstract: Pooled statistics of Transformer weights obscure how magnitude is distributed across functional channels, while individual weights are too numerous to compare directly. We study the mesoscopic level between them: row and column scale fields, the med…
Read the full story at arXiv stat.ML ↗
Timeline · 1 report
- 2026-10-01 04:00 · arXiv stat.ML
A Mesoscopic View of Transformer Weights Through Row and Column Scale Fields