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

Language Modeling on WikiText (test)

5.49Perplexity

Full

Updated 4d ago

Evaluation Results

Method	Links
Full 2025.12		5.49	-	-	-	-
TEAL 2025.12		5.52	-	-	-	-
SPON 2025.12		5.58	-	-	-	-
SPON 2025.12		5.86	-	-	-	-
TEAL 2025.12		6	-	-	-	-
Wanda 2025.12		6.26	-	-	-	-
SPON 2025.12		6.51	-	-	-	-
Full 2025.12		6.75	-	-	-	-
TEAL 2025.12		6.88	-	-	-	-
TEAL 2025.12		6.9	-	-	-	-
SPON 2025.12		6.92	-	-	-	-
SPON 2025.12		7.83	-	-	-	-
TEAL 2025.12		8.34	-	-	-	-
SPON 2025.12		8.94	-	-	-	-
Full 2025.12		8.99	-	-	-	-
TEAL 2025.12		9.04	-	-	-	-
SPON 2025.12		9.26	-	-	-	-
SPON 2025.12		9.63	-	-	-	-
Wanda 2025.12		9.66	-	-	-	-
TEAL 2025.12		9.75	-	-	-	-
Wanda 2025.12		10.41	-	-	-	-
SPON 2025.12		10.42	-	-	-	-
TEAL 2025.12		11.38	-	-	-	-
TEAL 2025.12		11.62	-	-	-	-
Full 2025.12		12.14	-	-	-	-
SPON 2025.12		15.43	-	-	-	-
TEAL 2025.12		17.03	-	-	-	-
GPT-2 (MLE) 2023.05		18.24	90	0.75	56.7	-
GPT-2 (MLE) 2023.05		20.43	90	0.73	57.3	-
GPT-2 (MIXCE) 2023.05		23.44	88	0.95	57.8	-
AR 2024.10		25.75	-	-	-	-
GPT-2 (MIXCE) 2023.05		25.92	88	0.95	58.4	-
GPT-2 (MLE) 2023.05		26.98	91	0.67	55.6	-
DenseFormer 2024.10		28	-	-	-	-
EDLM-coAR 2024.10		28.31	-	-	-	-
EDLM-AR 2024.10		29.24	-	-	-	-
Wanda 2025.12		29.68	-	-	-	-
EDLM-NCE 2024.10		30.77	-	-	-	-
Learnable ResFormer plus 2024.10		31.4	-	-	-	-
Learnable ResFormer 2024.10		32.5	-	-	-	-
MLDM 2024.10		32.83	-	-	-	-
Identity ResFormer 2024.10		32.9	-	-	-	-
Transformer 2024.10		33.4	-	-	-	-
SEDD 2024.10		34.28	-	-	-	-
Donor 2026.01		34.56	-	-	-	-
GPT-2 (MIXCE) 2023.05		35.04	87	0.93	56.7	-
NeuTRENO 2024.10		36.3	-	-	-	-
Fixed Neuron Pruning 2025.12		62.36	-	-	-	-
LoRA 2026.01		98.37	-	-	-	-
IA3 2026.01		214.99	-	-	-	-
Top-K 2026.01		365.83	-	-	-	-
Zero-shot 2026.01		397.78	-	-	-	-
Human 2023.05		-	89	1	62.8	-
Human 2023.05		-	89	1	62.8	-
MLE 2023.05		-	89	0.93	58.4	-
MIXCE 2023.05		-	87	0.95	56.8	-
MLE 2023.05		-	88	0.95	60.2	-
MIXCE 2023.05		-	87	0.96	59	-
MLE 2023.05		-	89	0.96	59.4	-
MIXCE 2023.05		-	87	0.97	58	-