List your company

Question Answering on ARC-C (Accuracy, Avg Tokens)

0.88AccuracyDeepSeek-R1
0.76040.791450.82250.85355Mar 10, 2026

Evaluation results

MethodLinks
DeepSeek-R1Setting=ARM only2026.030.88504
Qwen3-1.7BSetting=ARM only2026.030.86397
LLaDA-8B → Llama-3.2-3BSetting=text-space2026.030.8254
Latent-DARM (LLaDA-8B → Llama-3.2-3B)Setting=latent-space, Plan tokens=2562026.030.812
Llama-3.2-3BSetting=ARM only2026.030.7954
Latent-DARM (LLaDA-8B → Llama-3.2-3B)Setting=latent-space, Plan tokens=642026.030.7852
Latent-DARM (LLaDA-8B → Llama-3.2-3B)Setting=latent-space, Plan tokens=1282026.030.7652