AI.info

Yoonjun Cho

Explore Yoonjun Cho on AI.info.

• Preserve-Then-Quantize: Balancing Rank Budgets for Quantization Error Reconstruction in LLMs