Linear Regression In-context Learning on Garg Linear Regression d=25
0.991R^2Transformer
Evaluation results
| Method | Links | |
|---|---|---|
| TransformerTraining Steps=50k2026.07 | 0.991 | |
| TransformerCompute=50k2026.07 | 0.991 | |
| DendriCLTraining Steps=50k2026.07 | 0.809 | |
| DendriCLCompute=50k2026.07 | 0.809 | |
| DendriCL2026.07 | 0.791 | |
| DendriCLCompute=10k2026.07 | 0.791 | |
| SpikformerTraining Steps=50k2026.07 | 0.656 | |
| SpikformerCompute=50k2026.07 | 0.656 | |
| Pure LIFTraining Steps=50k2026.07 | 0.588 | |
| Transformer2026.07 | 0.58 | |
| TransformerCompute=10k2026.07 | 0.58 | |
| Pure LIF2026.07 | 0.008 | |
| Pure LIFCompute=10k2026.07 | 0.008 |