The first launch of Energy-Based Reasoners.

Energy-Based Reasoners
Launching todayA 1.8M-parameter model that cracks the hardest Sudoku
Meet the team
Hunted by Ilyas. The makers are not on 25launches yet.

A 1.8M-parameter model that cracks the hardest Sudoku
0 followers
Artificial IntelligenceOpen Source
Energy-Based Reasoners (EBRs) are a research paper with code from Duke, UIUC and Harvard. Instead of looping a network a fixed number of times, an EBR learns an energy score for its current reasoning state, refines it by gradient descent, stops when it settles and keeps the lowest-energy answer. With 1.8M parameters it solves 99.4% of Sudoku-Extreme in one try and 99.9% with a few samples, using 345x less inference compute than the next-best model. Code is MIT; training needs a CUDA GPU.
The first launch of Energy-Based Reasoners.

A 1.8M-parameter model that cracks the hardest Sudoku
Hunted by Ilyas. The makers are not on 25launches yet.