List your company

Verdict Prediction on ConsistencyCheck

85.8AccuracyDeepSeek-V3.2
39.5251.53563.5575.565Jul 6, 2026

Evaluation results

MethodLinks
DeepSeek-V3.22026.0785.865.269.867.4
GPT-5-mini2026.078481.667.774
Claude-Sonnet-42026.0783.378.968.873.5
Claude-Sonnet-4.62026.0782.992.44964
GPT-5.3-Codex2026.0782.272.672.672.6
GPT-4.12026.0777.472.552.961.2
LeanScorer2026.0776.873.249.158.8
GPT-5.22026.077155.187.267.5
FormalRx-8BSize=8B2026.0770.254.865.459.6
FormalRx-4BSize=4B2026.0767.250.968.958.5
Qwen3-8BSize=8B2026.0766.850.66456.5
FormalRx-1.7BSize=1.7B2026.0765.3497358.6
Qwen3-4BSize=4B2026.0758.34373.454.2
Qwen3-1.7BSize=1.7B2026.0753.639.369.950.3
BLEU (Best F1)2026.0741.335.994.552