List your company

Simulatability on Tabular Datasets Aggregate (test)

74.63Accuracy (w/o explanation)Gemini 3
63.366866.290969.21572.1391Feb 2, 2026

Evaluation results

MethodLinks
Gemini 3Variant=Flash2026.0274.6382.447.81—30.79—
Claude 4.5Variant=Opus2026.027481.387.38—28.38—
Qwen 3Model Size=14B2026.0272.8982.089.19—33.9—
Claude 4.5Variant=Sonnet2026.0272.5381.819.28—33.78—
GPT-5.22026.0272.582.059.55—34.72—
GPT-5Variant=mini2026.0272.3881.849.46—34.25—
Qwen 3Model Size=8B2026.0272.1580.38.14—29.24—
Claude 4.5Variant=Haiku2026.0271.9879.497.51—26.8—
Gemma 3Model Size=12B2026.0271.6481.219.57—33.74—
Gemini 3Variant=Pro2026.0271.6279.688.06—28.38—
Qwen 3Model Size=32B2026.0271.1881.4410.27—35.61—
Gemma 3Model Size=27B2026.0271.0181.1410.12—34.92—
Gemma 3Model Size=4B2026.0270.3781.1910.82—36.51—
GPT-5Variant=nano2026.0269.4879.4910.02—32.81—
Qwen 3Model Size=4B2026.0269.2576.697.44—24.19—
Qwen 3Model Size=1.7B2026.0266.1771.55.34—15.77—
Qwen 3Model Size=0.6B2026.0265.7369.513.78—11.03—
Gemma 3Model Size=1B2026.0263.870.146.34—17.51—