arxiv:2510.04212
Haiquan Qiu
huggingaaaaa
AI & ML interests
None yet
Recent Activity
authored
a paper
16 days ago
Spectral Alignment as Predictor of Loss Explosion in Neural Network
Training
authored
a paper
20 days ago
Why Low-Precision Transformer Training Fails: An Analysis on Flash
Attention
upvoted
a
paper
20 days ago
Why Low-Precision Transformer Training Fails: An Analysis on Flash
Attention
Organizations
None yet