List your company

Mathematical Reasoning on MATH (test) (Accuracy + Throughput/Speed)

30.1AccuracyLLaDA
25.31626.55827.829.042Jan 28, 2026

Evaluation results

MethodLinks
LLaDAModel=LLaDA-Base, #Shots=4, Max generation length=2562026.0130.15.061
Fast-dLLM (Prefix-Cache)Model=LLaDA-Base, #Shots=4, Max generation length=256, block size=32, parallel decoding=disabled2026.0128.515.393.04
DKV-CacheModel=LLaDA-Base, #Shots=4, Max generation length=256, cache refresh interval=8, block size=322026.0127.88.791.74
Window-DiffusionModel=LLaDA-Base, #Shots=4, Max generation length=256, external window length=64, internal window length=16, cache refresh interval=32, early stopping=disabled2026.0126.225.935.12
Fast-dLLM (Dual-Cache)Model=LLaDA-Base, #Shots=4, Max generation length=256, block size=32, parallel decoding=disabled2026.0125.518.373.63