List your company

Question Answering on WMDP Biology

64.5Default ScoreBase model
24.9835.2445.555.76Jan 31, 2025

Evaluation results

MethodLinks
Base modelRefinement=Original, Attack forget-set source=WMDP-Cyber2025.0164.5————
Base modelRefinement=Original, Attack forget-set source=WMDP-Biology2025.0164.5————
DPO+MSERefinement=w/ RNA, Attack forget-set source=WMDP-Cyber2025.0133.3—49.649.438
DPO+MSERefinement=w/ RNA, Attack forget-set source=WMDP-Biology2025.0133.333.251.749.652.3
SimNPO+MSERefinement=w/ RNA, Attack forget-set source=WMDP-Cyber2025.0131.5—60.637.960.6
SimNPO+MSERefinement=w/ RNA, Attack forget-set source=WMDP-Biology2025.0131.532.660.537.557
Adaptive RMURefinement=w/ RNA, Attack forget-set source=WMDP-Cyber2025.0131.3—55.934.257.3
RSVRefinement=w/ RNA, Attack forget-set source=WMDP-Cyber2025.0131.3—48.540.136.9
Adaptive RMURefinement=w/ RNA, Attack forget-set source=WMDP-Biology2025.0131.330.258.834.656.1
RSVRefinement=w/ RNA, Attack forget-set source=WMDP-Biology2025.0131.330.159.837.658.1
NPO+MSERefinement=w/ RNA, Attack forget-set source=WMDP-Cyber2025.0130.6—6134.144.9
NPO+MSERefinement=w/ RNA, Attack forget-set source=WMDP-Biology2025.0130.629.859.732.455.8
RMURefinement=w/ RNA, Attack forget-set source=WMDP-Cyber2025.0130.5—58.130.750
RMURefinement=w/ RNA, Attack forget-set source=WMDP-Biology2025.0130.529.159.442.856.8
NPO+KLRefinement=w/ RNA, Attack forget-set source=WMDP-Cyber2025.0129.6—62.736.158.5
SimNPO+KLRefinement=w/ RNA, Attack forget-set source=WMDP-Cyber2025.0129.6—63.133.760
NPO+KLRefinement=w/ RNA, Attack forget-set source=WMDP-Biology2025.0129.629.161.749.756.3
SimNPO+KLRefinement=w/ RNA, Attack forget-set source=WMDP-Biology2025.0129.630.462.539.257.3
Adaptive RMURefinement=Original, Attack forget-set source=WMDP-Cyber2025.0129.4—61.441.353.1
DPO+KLRefinement=w/ RNA, Attack forget-set source=WMDP-Cyber2025.0129.4—59.443.745.6
Adaptive RMURefinement=Original, Attack forget-set source=WMDP-Biology2025.0129.427.964.843.957
DPO+KLRefinement=w/ RNA, Attack forget-set source=WMDP-Biology2025.0129.428.459.636.155
DPO+KLRefinement=Original, Attack forget-set source=WMDP-Cyber2025.0129.1—60.534.253.7
DPO+KLRefinement=Original, Attack forget-set source=WMDP-Biology2025.0129.127.959.743.453.9
NPO+KLRefinement=Original, Attack forget-set source=WMDP-Cyber2025.0128.8—55.534.549.8
NPO+KLRefinement=Original, Attack forget-set source=WMDP-Biology2025.0128.827.762.432.256.4
RMURefinement=Original, Attack forget-set source=WMDP-Cyber2025.0128—61.227.453.9
DPO+MSERefinement=Original, Attack forget-set source=WMDP-Cyber2025.0128—46.432.544.9
RMURefinement=Original, Attack forget-set source=WMDP-Biology2025.012827.76430.957.7
DPO+MSERefinement=Original, Attack forget-set source=WMDP-Biology2025.012827.355.133.555.1
RSVRefinement=Original, Attack forget-set source=WMDP-Cyber2025.0127.8—50.432.637.2
RSVRefinement=Original, Attack forget-set source=WMDP-Biology2025.0127.827.261.93757.8
SimNPO+MSERefinement=Original, Attack forget-set source=WMDP-Cyber2025.0127.6—61.727.254.2
SimNPO+MSERefinement=Original, Attack forget-set source=WMDP-Biology2025.0127.627.261.232.355.1
NPO+MSERefinement=Original, Attack forget-set source=WMDP-Cyber2025.0127—60.53152.7
NPO+MSERefinement=Original, Attack forget-set source=WMDP-Biology2025.012728.36152.955.9
SimNPO+KLRefinement=Original, Attack forget-set source=WMDP-Cyber2025.0126.5—61.632.645.3
SimNPO+KLRefinement=Original, Attack forget-set source=WMDP-Biology2025.0126.52762.229.755.7