List your company

Mathematical Reasoning on GSM8K (ASR, Acc)

99.2ASRNo Def.
-0.1225.66551.4577.235Apr 27, 2026

Evaluation results

MethodLinks
No Def.Attack=BadEdit2026.0499.282.5
No Def.Attack=BadChain2026.0498.481.8
No Def.Attack=EmbedX2026.0497.782.3
No Def.Attack=VPI2026.0495.882.5
DemoDefenseAttack=BadChain2026.0428.280
FABEAttack=BadChain2026.0424.976.5
FABEAttack=EmbedX2026.0422.377.1
LLMBDAttack=BadChain2026.0422.280.5
LLMBDAttack=EmbedX2026.0419.473.7
FABEAttack=BadEdit2026.0418.677.3
LLMBDAttack=BadEdit2026.0418.174.2
DemoDefenseAttack=EmbedX2026.041880.7
DemoDefenseAttack=BadEdit2026.0416.381.5
CROWAttack=BadChain2026.041679.7
FABEAttack=VPI2026.0415.877.3
TIGSAttack=EmbedX2026.041480.1
BEEARAttack=BadChain2026.0413.578.1
CROWAttack=EmbedX2026.0412.479.9
LLMBDAttack=VPI2026.0412.174.2
W2SDefenseAttack=BadChain2026.0411.680.1
W2SDefenseAttack=EmbedX2026.0411.681.1
BEEARAttack=EmbedX2026.0410.779.3
TIGSAttack=BadChain2026.0410.580.5
TIGSAttack=BadEdit2026.049.680.1
CROWAttack=VPI2026.04880.7
CROWAttack=BadEdit2026.046.580.7
BEEARAttack=VPI2026.046.280.8
W2SDefenseAttack=BadEdit2026.045.280.4
BEEARAttack=BadEdit2026.044.980.8
W2SDefenseAttack=VPI2026.044.881.7
TIGSAttack=VPI2026.044.280.6
DemoDefenseAttack=VPI2026.043.778.7