List your company

ResearchFormerly WizWand

Natural Language Inference on AmericasNLI (test)

58.9Accuracy (aym)BISTIL-TF

Evaluation results

Method	Links
BISTIL-TF2023.06		58.9	45.7	46.4	62.9	44.3	50.8	44	58.7	47.2	43.1	50.2	-1.1	—
LTSFT2023.06		58.1	44.4	47.9	63.5	42.8	52.4	48.5	62	50.3	43.3	51.3	—	—
BISTIL-TF2023.06		57.7	43.6	48.1	60.9	41.3	51.4	42.6	59.9	45.5	40.3	49.1	-2.2	—
UniBridge2024.06		52.13	36.8	40.26	—	35.86	46.88	42.38	59.86	35.6	36.4	44.58	—	59.59
MAD-X2024.06		39.46	36.8	38.93	—	35.86	40.78	33.42	37.46	39.06	36.53	37.8	—	39.73
XLM-R2024.06		36.26	38.53	36.4	—	37.33	39.43	36.89	37.6	35.86	34.66	37.03	—	37.33
UniBridge2024.06		35.73	33.33	33.33	—	34.4	36.31	33.42	36.66	34.4	34.4	34.94	—	37.46
mBERT2024.06		33.33	33.33	33.33	—	33.33	33.33	33.2	33.28	33.33	33.33	33.31	—	33.33
MAD-X2024.06		33.06	33.33	34.4	—	34	33.33	33.42	33.73	32.93	33.2	33.54	—	33.46