Mathematical Reasoning on GSM8K (Base Metrics Set)
78.8Base AccuracyLlama-3-8B-Instruct
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Llama-3-8B-InstructSetting=Base2026.04 | 78.8 | — | — | — | — | — | — | — | — | — | — | |
| IRISBackbone=Llama-3-8B-Instruct2026.04 | — | 72.4 | 79.8 | 77.9 | 70.4 | 73.5 | 75.4 | 76.7 | 79.7 | 77.8 | 69.3 |