Sudoku Solving on Sudoku (n=5) 25x25 (val)
72AccuracyMPLM
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MPLMReasoning Paradigm=Message Passing2026.07 | 72 | 1,017.263 | |
| GPT-5 Proqueries=20, API=batched2026.07 | 20 | — | |
| DeepSeek R1queries=202026.07 | 0 | 431.751 |
Loading the SOTA2 catalog…