Symbolic Reasoning on Maze Hard (test)
87.7AccuracyRIMformer
Evaluation Results
| Method | Links | |
|---|---|---|
| RIMformerReweighter=expressive2026.03 | 87.7 | |
| RIMAReweighter=expressive2026.03 | 87 | |
| SimRIM (aka TRM)Reweighter=identity2026.03 | 85.3 |
Loading the SOTA2 catalog…
| Method | Links | |
|---|---|---|
| RIMformerReweighter=expressive2026.03 | 87.7 | |
| RIMAReweighter=expressive2026.03 | 87 | |
| SimRIM (aka TRM)Reweighter=identity2026.03 | 85.3 |