List your company

Multi-task Language Understanding Performance on MMLU (TPS & Speedup)

17.3TPSCreditDecoding
4.4567.790511.12514.4595Oct 7, 2025

Evaluation results

MethodLinks
CreditDecodingBackbone=LLaDA-8B-Instruct, Gen Length=256, Block Size=642025.1017.324938.4——
Fast-dLLMBackbone=LLaDA-8B-Instruct, Gen Length=256, Block Size=642025.1012.49————
BaselineBackbone=LLaDA-8B-Instruct, Gen Length=256, Block Size=642025.104.95————
BaselineBackbone=LLaMA 3.1 8B, Shots=5-shot2025.10———59.2—
Best ModelBackbone=LLaMA 3.1 8B, Shots=5-shot, #D=42025.10———62.381.14
BSBABackbone=LLaMA 3.1 8B, Shots=5-shot, #D=72025.10———59.571.26