tegmark
@ejio143612
1/9 🚨 New Paper Alert: Cross-Entropy Loss is NOT What You Need! 🚨 We introduce harmonic loss as alternative to the standard CE loss for training neural networks and LLMs! Harmonic loss achieves 🛠?significantly better interpretability, ?faster convergence, and ?less grokking! https://t.co/DratQQV8q0
0 reply
0 recast
0 reaction