Our new X account is live! Follow @wizwand_team for updates

Machine Unlearning on TOFU

1Forget Quality (FQ)

Retrained LLM

Updated 4d ago

Evaluation Results

Method
Retrained LLM 2026.02	1	0.6267	0.408	0.9833
Retrained LLM 2026.02	1	0.5233	0.4272	0.9269
Residual Knowledge risk score 2026.01	0.83	0.49	-	-
Predictive Entropy 2026.01	0.77	0.48	-	-
No Abstention 2026.01	0.72	0.51	-	-
MOX 2026.02	0.068	0.6528	0.441	0.9802
MOX 2026.02	0.0677	0.6412	0.4788	0.971
MOX 2026.02	0.0625	0.6504	0.4697	0.9653
MOX 2026.02	0.0598	0.551	0.312	0.8988
MOX 2026.02	0.0582	0.5219	0.3138	0.881
PO 2026.02	0.0541	0.6308	0.364	0.8811
LLMU 2026.02	0.0541	0.6337	0.448	0.8865
DPO 2026.02	0.0541	0.6359	0.586	0.8852
DPO 2026.02	0.0521	0.5125	0.3437	0.7349
RMU 2026.02	0.0387	0.5528	0.515	0.8488
MOX 2026.02	0.0364	0.5012	0.433	0.8928
MOX 2026.02	0.034	0.515	0.3436	0.8562
MOX 2026.02	0.0328	0.5012	0.3366	0.8858
MOX 2026.02	0.0319	0.642	0.4658	0.9016
SimNPO 2026.02	0.0315	0.5665	0.5135	0.8514
PO 2026.02	0.0286	0.5127	0.317	0.7468
LLMU 2026.02	0.0286	0.511	0.3058	0.727
GAD 2026.02	0.0268	0.632	0.4773	0.8912
MOX 2026.02	0.0256	0.641	0.4555	0.9701
AltPO 2026.02	0.0231	0.5312	0.52	0.8288
GAD 2026.02	0.0215	0.511	0.4996	0.8496
GA 2026.02	0.0213	0.5069	0.5114	0.8048
RMU 2026.02	0.0211	0.6378	0.4645	0.9032
MOX 2026.02	0.0182	0.6358	0.4732	0.9788
MOX 2026.02	0.018	0.5026	0.4366	0.912
SimNPO 2026.02	0.0172	0.645	0.4601	0.9201
KL 2026.02	0.0168	0.63	0.5281	0.9398
MOX 2026.02	0.0163	0.5002	0.4512	0.92
TV 2026.02	0.0156	0.5012	0.4366	0.881
MOX 2026.02	0.0146	0.6305	0.4812	0.981
GA 2026.02	0.0143	0.6333	0.4862	0.9008
AltPO 2026.02	0.012	0.6432	0.465	0.8953
KL 2026.02	0.012	0.5047	0.5059	0.8109
TV 2026.02	0.0069	0.634	0.4512	0.981
NPO 2026.02	0.0068	0.6321	0.4632	0.895
NPO 2026.02	0.003	0.5057	0.5196	0.8
Original LLM 2026.02	0.0013	0.5184	0.9607	0.9199
Original LLM 2026.02	0	0.6346	0.9851	0.9833