List your company

Continual Learning on KQA → Math → Code → MQA Task-level stream

51.67AA ScoreMasFACT
30.006835.630941.25546.8791May 17, 2026

Evaluation results

MethodLinks
MasFACTReplay-based=false2026.0551.672.07
MasFACTReplay-based=false2026.0551.672.07
MasRouterReplay-based=true2026.0549.555.02
MasRouterReplay-based=true2026.0549.555.02
ARGDesignerReplay-based=true2026.0549.134.83
ARGDesignerReplay-based=true2026.0549.134.83
MasRouterReplay-based=false2026.0545.57.48
MasRouterReplay-based=false2026.0545.57.48
GDesignerReplay-based=true2026.0544.215.15
GDesignerReplay-based=true2026.0544.215.15
ARGDesignerReplay-based=false2026.0542.186.14
ARGDesignerReplay-based=false2026.0542.186.14
GDesignerReplay-based=false2026.0541.816.95
GDesignerReplay-based=false2026.0541.816.95
AgentDropoutReplay-based=true2026.0540.617.48
AgentDropoutReplay-based=true2026.0540.617.48
AFLOWReplay-based=true2026.0540.394.12
AFLOWReplay-based=true2026.0540.394.12
AFLOWReplay-based=false2026.0537.766.45
AFLOWReplay-based=false2026.0537.766.45
SC(CoT)-ICLReplay-based=false2026.0537.475.63
SC(CoT)-ICLReplay-based=false2026.0537.475.63
AgentDropoutReplay-based=false2026.0536.728.12
AgentDropoutReplay-based=false2026.0536.728.12
CompleteReplay-based=false2026.0536.06—
Complete GraphReplay-based=false2026.0536.06—
CoT-ICLReplay-based=false2026.0534.95.64
CoT-ICLReplay-based=false2026.0534.95.64
TreeReplay-based=false2026.0534.17—
TreeReplay-based=false2026.0534.17—
ChainReplay-based=false2026.0532.24—
ChainReplay-based=false2026.0532.24—
VanillaReplay-based=false2026.0530.84—
VanillaReplay-based=false2026.0530.84—