List your company

ResearchFormerly WizWand

Big-Bench Hard Reasoning

Benchmarks

DatasetSOTA methodMetricTrendResultsLast updated

69.16Accuracy

14

Apr 7, 2026

BBH n=2395 (test)

100Activation

11

Jun 30, 2026

Dream Base + dLLM-Cache

29.61Throughput (TPS)

4

Jun 4, 2026