Continual Learning on Code Domain-level stream
62.09AAMasRouter
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MasRouterReplay-based=true2026.05 | 62.09 | 3.46 | |
| MasRouterReplay-based=true2026.05 | 62.09 | 3.46 | |
| MasFACTReplay-based=false2026.05 | 61.14 | -2.03 | |
| MasFACTReplay-based=false2026.05 | 61.14 | -2.03 | |
| ARGDesignerReplay-based=true2026.05 | 58.41 | 2.76 | |
| ARGDesignerReplay-based=true2026.05 | 58.41 | 2.76 | |
| MasRouterReplay-based=false2026.05 | 58.03 | 4.18 | |
| MasRouterReplay-based=false2026.05 | 58.03 | 4.18 | |
| GDesignerReplay-based=true2026.05 | 57.13 | 5.04 | |
| GDesignerReplay-based=true2026.05 | 57.13 | 5.04 | |
| AFLOWReplay-based=true2026.05 | 54.77 | 4.22 | |
| AFLOWReplay-based=true2026.05 | 54.77 | 4.22 | |
| GDesignerReplay-based=false2026.05 | 54.46 | 4.89 | |
| GDesignerReplay-based=false2026.05 | 54.46 | 4.89 | |
| AgentDropoutReplay-based=true2026.05 | 53.58 | 2.28 | |
| AgentDropoutReplay-based=true2026.05 | 53.58 | 2.28 | |
| ARGDesignerReplay-based=false2026.05 | 52.47 | 4.09 | |
| ARGDesignerReplay-based=false2026.05 | 52.47 | 4.09 | |
| SC(CoT)-ICLReplay-based=false2026.05 | 50.46 | 3.4 | |
| SC(CoT)-ICLReplay-based=false2026.05 | 50.46 | 3.4 | |
| AgentDropoutReplay-based=false2026.05 | 48.91 | 4.03 | |
| AgentDropoutReplay-based=false2026.05 | 48.91 | 4.03 | |
| AFLOWReplay-based=false2026.05 | 48.82 | 6.17 | |
| AFLOWReplay-based=false2026.05 | 48.82 | 6.17 | |
| TreeReplay-based=false2026.05 | 47.12 | — | |
| TreeReplay-based=false2026.05 | 47.12 | — | |
| CoT-ICLReplay-based=false2026.05 | 46.74 | 3.29 | |
| CoT-ICLReplay-based=false2026.05 | 46.74 | 3.29 | |
| ChainReplay-based=false2026.05 | 45.64 | — | |
| ChainReplay-based=false2026.05 | 45.64 | — | |
| VanillaReplay-based=false2026.05 | 43.78 | — | |
| VanillaReplay-based=false2026.05 | 43.78 | — | |
| CompleteReplay-based=false2026.05 | 41.71 | — | |
| Complete GraphReplay-based=false2026.05 | 41.71 | — |