$XIAOMI-W (01810.HK)$ Industry insiders in large models have a few brief comments. Using RL (Reinforcement Learning) to optimize mathematical and programming abilities is standard practice, like fine-tuning for specific tasks. Other large models emphasize general applicability, while Xiaomi focuses on competing with mathematical datasets individually—if the results aren't up to par, that would be surprising. 360 actually provided similar results in the early stages (Light-R1).
In the short term, this news is indeed Bullish, as Xiaomi's stock price is clearly driven by sentiment (it's close to meme stocks); however, the long-term impact is limited since the technical barrier lies not in single-point optimization, but in overall capability.