List your company

General Reasoning on GPQA (Accuracy, Loss)

25.8AccuracyAdam
24.96825.18425.425.616Apr 10, 2026

Evaluation results

MethodLinks
AdamModel scale=1B2026.0425.82.28
NexusModel scale=3B2026.0425.82.047
NexusModel scale=1B2026.04252.261
AdamModel scale=3B2026.04252.062