Sudoku Solving on Sudoku (n=4) 16x16 (val)
92AccuracyMPLM
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MPLMReasoning Paradigm=Message Passing2026.07 | 92 | 117.684 | |
| GPT-5 Proqueries=20, API=batched2026.07 | 45 | — | |
| DeepSeek R1queries=202026.07 | 0 | 597.939 |
Loading the SOTA2 catalog…