List your company

General Reasoning on GPQA (Accuracy)

41.4GPQA AccuracySDAR-8B-Chat
26.184830.134934.08538.0351Apr 5, 2026

Evaluation results

MethodLinks
SDAR-8B-Chatgeneration length=optimal, block length=optimal, denoising steps=optimal2026.0441.4
SDAR-30B-A3Bgeneration length=optimal, block length=optimal, denoising steps=optimal2026.0436.36
LLaDA2.0-minigeneration length=optimal, block length=optimal, denoising steps=optimal2026.0434.34
LLaDA2.1-minigeneration length=optimal, block length=optimal, denoising steps=optimal2026.0434.34
LLaDA-8B-Instructgeneration length=optimal, block length=optimal, denoising steps=optimal2026.0431.82
Dream-7B-Instructgeneration length=optimal, block length=optimal, denoising steps=optimal2026.0426.77