Heykuki News
Top
New
Best
Ask
Show
Jobs
Toggle theme
Login
Top
New
Best
Ask
Show
Jobs
A 27M-param model that solves hard Sudoku/mazes where LLMs fail, without CoT
github.com/sapientinc
10 points
mingli_yuan
10 months ago
1 comment
Loading...
A 27M-param model that solves hard Sudoku/mazes where LLMs fail, without CoT | Heykuki News