Showing cs.LGShow all
2 papers · 1 filter
cs.LG2025
Chebyshev Moment Regularization (CMR): Condition-Number Control with Moment Shaping
Jinwoo Baek
We introduce \textbf{Chebyshev Moment Regularization (CMR)}, a simple, architecture-agnostic loss that directly optimizes layer spectra. CMR jointly controls spectral edges via a l…
cs.LG2025
Numerical Fragility in Transformers: A Layer-wise Theory for Risk Estimation and Selective Stabilization
Jinwoo Baek
Low-precision execution can induce substantial forward discrepancies in Transformers even for fixed weights and input, yet these discrepancies are usually monitored only at the out…