This empirical study dissects the serving dynamics of reasoning large language models, revealing critical performance trade-offs and optimization strategies ...
Level: advanced
By Qi Li and 8 other authors
Category: research