AI.info

Haozhe Hu

Explore Haozhe Hu on AI.info.

• WIDE: Boosting Adaptive LLM Inference via Token-level Dynamic Width Pruning