Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Adversarial Generation on WOMD

65.57Attack Success Rate

Rule

10.335624.675339.01553.3547Sep 24, 2025Nov 2, 2025Dec 11, 2025Jan 19, 2026Feb 27, 2026Apr 7, 2026May 17, 2026
Updated 14d ago

Evaluation Results

MethodLinks
2026.05
65.572.761-113.71.8036.14810.8513.470.336
2025.09
60.542.7634.83336.9471.9736.8564.4668.4470.309
2025.09
43.481.74610.2753.2073.3458.6961.5957.550.231
2026.05
391.87-3.0020.0202.2170.8380.202
2025.09
33.70.9425.7852.7193.56211.2331.6957.3680.276
2025.09
33.511.7791.4332.480.791.092.0988.070.185
2026.05
31.40.752-2.6843.0311.981.5638.2670.267
2026.05
30.331.319-3.0392.6236.9671.5398.8770.187
2026.05
28.421.4-2.4960.7921.3662.0988.1140.188
2025.09
20.981.1830.4982.0680.8170.9541.5228.4580.08
2025.09
20.440.9634.5282.7972.56110.3541.756.1730.27
2026.05
19.40.992-2.7792.4599.9731.7496.1870.269
2026.05
19.141.148-2.5963.1145.8570.983259.10.097
2025.09
16.481.1462.4512.5663.0975.8240.955257.6190.097
2025.09
16.390.4395.09211.253.72212.643.7168.2230.157
2025.09
13.90.8640.33220.7080.9541.4619.3180.055
2026.05
12.460.667-15.473.50711.454.6628.070.152