P
PaperPicks
Conferences
Atli Kosson
4 papers at tracked venues · 4 at CORE A* · active 2024–2024
DBLP profile ↗
ORCID search ↗
Venues
NeurIPS
×2
AAAI
×1
ICML
×1
Frequent coauthors
Alexander Hägele
DBLP profile ↗
ORCID search ↗
×1
Papers
Analyzing & Reducing the Need for Learning Rate Warmup in GPT Training
NeurIPS 2024
·
Atli Kosson
Ghost Noise for Regularizing Deep Neural Networks
AAAI 2024
·
Atli Kosson
Rotational Equilibrium: How Weight Decay Balances Learning Across Neural Networks
ICML 2024
·
Atli Kosson
Scaling Laws and Compute-Optimal Training Beyond Fixed Training Durations
NeurIPS 2024
·
Alexander Hägele
DBLP profile ↗
ORCID search ↗