Heykuki News
Top
New
Best
Ask
Show
Jobs
Toggle theme
Login
Top
New
Best
Ask
Show
Jobs
Show HN: Adaptive-K – Cut MoE inference costs 30-50% with entropy-guided routing
github.com/Gabrobals
1 point
Gabrielebalsamo
5 months ago
No comment yet
Show HN: Adaptive-K – Cut MoE inference costs 30-50% with entropy-guided routing | Heykuki News