Mathematical Reasoning on GSM8K (Normalized Performance)
1.026Normalized performanceIso-C
Evaluation Results
| Method | Links | |
|---|---|---|
| Iso-CCategory=State-of-the-Art (+Many-shot), Backbone=Llama-3.2-3B2026.06 | 1.026 | |
| FedMergeCategory=Iterative / FL, Backbone=Llama-3.2-3B2026.06 | 0.923 | |
| ColD FusionCategory=Iterative / FL, Backbone=Llama-3.2-3B2026.06 | 0.923 | |
| TIESCategory=Conventional Method (+Many-shot), Backbone=Llama-3.2-3B2026.06 | 0.923 | |
| Iso-CTSCategory=State-of-the-Art (+Many-shot), Backbone=Llama-3.2-3B2026.06 | 0.897 | |
| ConsensusTACategory=Conventional Method (+Many-shot), Backbone=Llama-3.2-3B2026.06 | 0.872 | |
| METISCategory=Ours, Backbone=Llama-3.2-3B2026.06 | 0.872 | |
| q-FedAvgCategory=Iterative / FL, Backbone=Llama-3.2-3B2026.06 | 0.846 | |
| DARECategory=Conventional Method (+Many-shot), Backbone=Llama-3.2-3B2026.06 | 0.846 | |
| TSV-MCategory=State-of-the-Art (+Many-shot), Backbone=Llama-3.2-3B2026.06 | 0.821 | |
| TULU 3Category=Data-Mixing, Backbone=Llama-3.2-3B2026.06 | 0.821 | |
| TA + Subspace BoostingCategory=State-of-the-Art (+Many-shot), Backbone=Llama-3.2-3B2026.06 | 0.744 | |
| Task ArithmeticCategory=Conventional Method (+Many-shot), Backbone=Llama-3.2-3B2026.06 | 0.744 |