List your company

Multi-task Language Understanding on MMLU STEM

89.1AccuracyMiMo-V2 Flash
22.43639.74357.0574.357May 19, 2025Jul 20, 2025Sep 20, 2025Nov 21, 2025Jan 22, 2026Mar 25, 2026May 26, 2026

Evaluation results

MethodLinks
MiMo-V2 FlashShots=5-shot, # Total Params=309B, # Active Params=15.1B2026.0589.1
Qwen3.5Shots=5-shot, # Total Params=35B, # Active Params=2.6B2026.0580.2
LAGUNA XS.2Shots=5-shot, # Total Params=33.4B, # Active Params=3B2026.0578.1
Nemotron-3-NanoShots=5-shot, # Total Params=31.6B, # Active Params=3.6B2026.0575.3
Gemma-4Shots=5-shot, # Total Params=25.2B, # Active Params=3.8B2026.0570.4
LMNetBackbone=Qwen2.5-0.5B2025.0546
Self-ConsistencyBackbone=Qwen2.5-0.5B, Samples=162025.0541.4
Self-RefineBackbone=Qwen2.5-0.5B, Passes=162025.0540.6
PromptBackbone=Qwen2.5-0.5B2025.0539.8
SFTBackbone=Qwen2.5-0.5B2025.0539.8
PolychromaticLMstage=SFT, shot=5-shot2026.0328.42
PolychromaticLMstage=Base, shot=5-shot2026.0325.28
Randomshot=5-shot2026.0325