List your company

Pointwise binary classification of political sentiment on D(NA) (full)

68.3Macro F1GPT 4o
65.38866.14466.967.656Feb 20, 2026

Evaluation results

MethodLinks
GPT 4o2026.0268.368.569.1
DeepSeek-V32026.026868.168.8
Ensemble 32026.0267.567.768.1
Grok 22026.0267.367.468.2
GPT 4.52026.0267.368.167.4
Gemini 2.5 Flash2026.0267.167.168.3
Llama 4 Maverick2026.0266.666.668
Claude 3.7 Sonnet2026.0266.667.366.6
Qwen 32026.0266.366.367.6
Ensemble 12026.02666668
Ensemble 22026.02666667.8
Gemini 1.5 Pro2026.026666.366.3
GPT 3.5 Turbo2026.0265.565.567.2