List your company

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

0.57Macro F1GPT 4o
0.297520.368260.4390.50974Feb 20, 2026

Evaluation results

MethodLinks
GPT 4o2026.020.570.570.575
DeepSeek-V32026.020.550.550.559
Ensemble 32026.020.5280.530.53
Grok 22026.020.5120.5150.53
GPT 3.5 Turbo2026.020.4970.510.534
Gemini 2.5 Flash2026.020.4810.50.527
GPT 4.52026.020.4480.550.509
Gemini 1.5 Pro2026.020.4440.460.447
Llama 4 Maverick2026.020.4410.4750.508
Qwen 32026.020.4350.460.489
Claude 3.7 Sonnet2026.020.4110.510.471
Ensemble 22026.020.3480.4450.493
Ensemble 12026.020.3080.4450.5