List your company

Grammaticality Judgment on RuBLiMP Russian 1.0

96.7AnaphoraHaiku (GPb+CT son)
86.92489.4629294.538Jun 2, 2025

Evaluation results

MethodLinks
Haiku (GPb+CT son)Backbone=Haiku, Prompting=Grammar Prompting + CoT (GPb+CT), Prompt Variant=son2025.0696.7741009410097.310094.6
gpt-4o (CT)Backbone=gpt-4o, Prompting=Chain-of-Thought (CT)2025.0696.787.396.795.986.79610094.2
gpt-4o (GPb+CT son)Backbone=gpt-4o, Prompting=Grammar Prompting + CoT (GPb+CT), Prompt Variant=son2025.0696.78810098.11009910097.4
gpt-4o (GPb+CT ol)Backbone=gpt-4o, Prompting=Grammar Prompting + CoT (GPb+CT), Prompt Variant=ol2025.0696.78410095.199.89899.396.1
Sonnet (base)Backbone=Sonnet, Prompting=base2025.0696909696.896.297.798.795.9
Haiku (GPb+CT ol)Backbone=Haiku, Prompting=Grammar Prompting + CoT (GPb+CT), Prompt Variant=ol2025.0695.380.799.394.899.3949894.5
gpt-4o (GPb ol)Backbone=gpt-4o, Prompting=Grammar Prompting (GPb), Prompt Variant=ol2025.0695.39010096.598.99710096.8
Sonnet (GPb son)Backbone=Sonnet, Prompting=Grammar Prompting (GPb), Prompt Variant=son2025.0695.389.310096.410099.799.397.2
Llama (GPb+CT)Prompting Strategy=Grammar Prompting + Chain-of-Thought, Backbone=o12025.0695.390.31009295.194.987.893.6
gpt-4o (base)Backbone=gpt-4o, Prompting=base2025.0694.788.794.794.190.796.398.794
Sonnet (GPb ol)Backbone=Sonnet, Prompting=Grammar Prompting (GPb), Prompt Variant=ol2025.0694.78810096.999.39998.796.7
Sonnet (GPb+CT son)Backbone=Sonnet, Prompting=Grammar Prompting + CoT (GPb+CT), Prompt Variant=son2025.0694.789.310097.71009910097.2
Sonnet (GPb+CT ol)Backbone=Sonnet, Prompting=Grammar Prompting + CoT (GPb+CT), Prompt Variant=ol2025.0694.786.710098.199.398.797.396.4
Sonnet (CT)Backbone=Sonnet, Prompting=Chain-of-Thought (CT)2025.069489.393.397.292.297.399.394.7
gpt-3.5 (GPb+CT ol)Backbone=gpt-3.5, Prompting=Grammar Prompting + CoT (GPb+CT), Prompt Variant=ol2025.0693.37010089.597.382.779.387.4
gpt-4o (GPb son)Backbone=gpt-4o, Prompting=Grammar Prompting (GPb), Prompt Variant=son2025.0693.390.710095.610098.310096.8
Llama (CT)Prompting Strategy=Chain-of-Thought (CT)2025.0693.390.5969479.98695.390.7
gpt-3.5 (GPb+CT son)Backbone=gpt-3.5, Prompting=Grammar Prompting + CoT (GPb+CT), Prompt Variant=son2025.0692.756.798.792.397.883.394.788
Haiku (GPb son)Backbone=Haiku, Prompting=Grammar Prompting (GPb), Prompt Variant=son2025.0692.773.399.388.597.696.310092.5
Haiku (GPb ol)Backbone=Haiku, Prompting=Grammar Prompting (GPb), Prompt Variant=ol2025.0692.779.310088.88888.787.389.3
Llama (Base)Prompting Strategy=Basic2025.0692.786.78891.975.683.79287.2
gpt-3.5 (base)Backbone=gpt-3.5, Prompting=base2025.069268.784.784.158.974.39079
Haiku (base)Backbone=Haiku, Prompting=base2025.0691.384.79090.163.688.398.786.7
Haiku (CT)Backbone=Haiku, Prompting=Chain-of-Thought (CT)2025.0690.784.790.79473.89110089.3
gpt-3.5 (GPb ol)Backbone=gpt-3.5, Prompting=Grammar Prompting (GPb), Prompt Variant=ol2025.0689.377.310084.586.46994.785.9
gpt-3.5 (CT)Backbone=gpt-3.5, Prompting=Chain-of-Thought (CT)2025.0688.777.36887.364.769.39077.9
gpt-3.5 (GPb son)Backbone=gpt-3.5, Prompting=Grammar Prompting (GPb), Prompt Variant=son2025.0688.76210084.885.87898.785.4
Llama (GPb)Prompting Strategy=Grammar Prompting (GPb), Backbone=o12025.0687.387.310089.591.18887.390.1