Velocity decoding on Public microelectrode benchmark Cross-Task (T-RT)
0.8778R^2RPNT (Full-SFT)
Evaluation results
| Method | Links | |
|---|---|---|
| RPNT (Full-SFT)Training regime=Pretrained, Fine-tuning strategy=Full dataset finetuning2026.01 | 0.8778 | |
| RPNT (FS-SFT)Training regime=Pretrained, Fine-tuning strategy=Few-shot finetuning2026.01 | 0.8515 | |
| RPNTTraining regime=From scratch2026.01 | 0.8356 | |
| o-POSSM-GRU (FT)Training regime=Pretrained, Fine-tuning strategy=Full model finetuning2026.01 | 0.7717 | |
| POSSM-GRU-SSTraining regime=From scratch2026.01 | 0.7687 | |
| o-POSSM-Mamba (FT)Training regime=Pretrained, Fine-tuning strategy=Full model finetuning2026.01 | 0.7621 | |
| POYO-1 (FT)Training regime=Pretrained, Fine-tuning strategy=Full model finetuning2026.01 | 0.7591 | |
| o-POSSM-S4D (FT)Training regime=Pretrained, Fine-tuning strategy=Full model finetuning2026.01 | 0.7584 | |
| POSSM-S4D-SSTraining regime=From scratch2026.01 | 0.7505 | |
| POSSM-Mamba-SSTraining regime=From scratch2026.01 | 0.7418 | |
| GRUTraining regime=From scratch2026.01 | 0.7279 | |
| POYO-SSTraining regime=From scratch2026.01 | 0.7156 | |
| S4DTraining regime=From scratch2026.01 | 0.7145 | |
| MLPTraining regime=From scratch2026.01 | 0.7007 | |
| MambaTraining regime=From scratch2026.01 | 0.6694 | |
| Wiener filterTraining regime=From scratch2026.01 | 0.5942 | |
| NDT-2 (FT)Training regime=Pretrained, Fine-tuning strategy=Full model finetuning2026.01 | 0.5903 |