List your company

Argumentation Quality Scoring on STEM Argumentation scoring dataset

68.7Exact MatchGPT-5.2
49.56454.53259.564.468Apr 25, 2026

Evaluation results

MethodLinks
GPT-5.2Prompt type=Calibrated Prompt2026.0468.789.90.440.030.710.708
GPT-5.1Prompt type=Calibrated Prompt2026.0467.689.90.450.060.6930.687
GPT-4o-miniPrompt type=Calibrated Prompt2026.0450.391.10.60.110.690.686