Continual Learning on Math Domain-level stream
47.91AAMasFACT
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MasFACTReplay-based=false2026.05 | 47.91 | 0.57 | |
| MasFACTReplay-based=false2026.05 | 47.91 | 0.57 | |
| MasRouterReplay-based=true2026.05 | 46.28 | 3.6 | |
| MasRouterReplay-based=true2026.05 | 46.28 | 3.6 | |
| GDesignerReplay-based=true2026.05 | 45.52 | 3.57 | |
| GDesignerReplay-based=true2026.05 | 45.52 | 3.57 | |
| MasRouterReplay-based=false2026.05 | 43.86 | 4.02 | |
| MasRouterReplay-based=false2026.05 | 43.86 | 4.02 | |
| GDesignerReplay-based=false2026.05 | 42.93 | 5.14 | |
| GDesignerReplay-based=false2026.05 | 42.93 | 5.14 | |
| AgentDropoutReplay-based=true2026.05 | 41.17 | 4.92 | |
| AgentDropoutReplay-based=true2026.05 | 41.17 | 4.92 | |
| ARGDesignerReplay-based=true2026.05 | 40.4 | 7.59 | |
| ARGDesignerReplay-based=true2026.05 | 40.4 | 7.59 | |
| AFLOWReplay-based=true2026.05 | 40.29 | 4.16 | |
| AFLOWReplay-based=true2026.05 | 40.29 | 4.16 | |
| AgentDropoutReplay-based=false2026.05 | 39.59 | 5.62 | |
| AgentDropoutReplay-based=false2026.05 | 39.59 | 5.62 | |
| AFLOWReplay-based=false2026.05 | 39.46 | 6.89 | |
| AFLOWReplay-based=false2026.05 | 39.46 | 6.89 | |
| ARGDesignerReplay-based=false2026.05 | 38.94 | 3.93 | |
| ARGDesignerReplay-based=false2026.05 | 38.94 | 3.93 | |
| CompleteReplay-based=false2026.05 | 35.5 | — | |
| Complete GraphReplay-based=false2026.05 | 35.5 | — | |
| TreeReplay-based=false2026.05 | 34.29 | — | |
| TreeReplay-based=false2026.05 | 34.29 | — | |
| SC(CoT)-ICLReplay-based=false2026.05 | 33.82 | 6.24 | |
| SC(CoT)-ICLReplay-based=false2026.05 | 33.82 | 6.24 | |
| ChainReplay-based=false2026.05 | 30.63 | — | |
| ChainReplay-based=false2026.05 | 30.63 | — | |
| CoT-ICLReplay-based=false2026.05 | 30.19 | 6.13 | |
| CoT-ICLReplay-based=false2026.05 | 30.19 | 6.13 | |
| VanillaReplay-based=false2026.05 | 29.28 | — | |
| VanillaReplay-based=false2026.05 | 29.28 | — |