AI.info

Smita Krishnaswamy

Explore Smita Krishnaswamy on AI.info.

• Dispersion Loss Counteracts Embedding Condensation and Improves Generalization in Small Language Models