List your company
ResearchFormerly WizWand

Multi-task Language Understanding on MMLU Pro (Subject Breakdown)

76.88MMLU Pro Engineering AccQwen3-Next-80B-A3B-Thinking
13.564830.002446.4462.8776Nov 25, 2025Nov 28, 2025Dec 2, 2025Dec 6, 2025Dec 10, 2025Dec 14, 2025Dec 18, 2025
Updated 5mo ago

Evaluation results

MethodLinks
Qwen3-Next-80B-A3B-ThinkingN. Params=80B, MoE=true, Reasoning=true2025.1276.8875.0672.95—
WMDBase Model=Qwen3-Next-80B-A3B-Thinking, N. Params=80B, MoE=true, Reasoning=true, Train Dataset=China-centric, Top-k Layers=8, Coeff=-0.25, Reposition=false2025.1274.9275.9273.35—
MDBase Model=Qwen3-Next-80B-A3B-Thinking, N. Params=80B, MoE=true, Reasoning=true, Train Dataset=Extended, Top-k Layers=4, Coeff=-1, Reposition=false2025.1274.7276.7772.14—
WRMDBase Model=Qwen3-Next-80B-A3B-Thinking, N. Params=80B, MoE=true, Reasoning=true, Train Dataset=China-centric, Top-k Layers=8, Coeff=-0.25, Reposition=false2025.1273.3776.1671.34—
MDBase Model=Qwen3-Next-80B-A3B-Thinking, N. Params=80B, MoE=true, Reasoning=true, Train Dataset=China-centric, Top-k Layers=4, Coeff=-0.5, Reposition=false2025.1272.5574.8269.34—
RMDBase Model=Qwen3-Next-80B-A3B-Thinking, N. Params=80B, MoE=true, Reasoning=true, Train Dataset=China-centric, Top-k Layers=4, Coeff=-0.75, Reposition=false2025.1271.5274.9469.74—
WMDBase Model=Qwen3-Next-80B-A3B-Thinking, N. Params=80B, MoE=true, Reasoning=true, Train Dataset=Extended, Top-k Layers=8, Coeff=-0.25, Reposition=false2025.1271.4175.6772.55—
RMDBase Model=Qwen3-Next-80B-A3B-Thinking, N. Params=80B, MoE=true, Reasoning=true, Train Dataset=Extended, Top-k Layers=4, Coeff=-2, Reposition=false2025.1270.7976.2872.14—
WRMDBase Model=Qwen3-Next-80B-A3B-Thinking, N. Params=80B, MoE=true, Reasoning=true, Train Dataset=Extended, Top-k Layers=16, Coeff=-0.15, Reposition=false2025.1267.775.3169.54—
ExpRAGLLM Backbone=Claude 3.7 Sonnet2025.1167—8085
Qwen3-Next-4B-ThinkingN. Params=4B, MoE=false, Reasoning=true2025.1266.7767.1155.31—
WRMDBase Model=Qwen3-Next-4B-Thinking, N. Params=4B, MoE=false, Reasoning=true, Train Dataset=Extended, Top-k Layers=2, Coeff=-2, Reposition=false2025.1266.1567.4857.11—
SelfRAGLLM Backbone=Claude 3.7 Sonnet2025.1165—7784
ReMemLLM Backbone=Claude 3.7 Sonnet2025.1165—8086
HistoryLLM Backbone=Claude 3.7 Sonnet2025.1164—7885
MemOSLLM Backbone=Claude 3.7 Sonnet2025.1164—7684
BaselineLLM Backbone=Claude 3.7 Sonnet2025.1163—7884
ReActLLM Backbone=Claude 3.7 Sonnet2025.1163—7684
AmemLLM Backbone=Claude 3.7 Sonnet2025.1163—7983
ExpRecentLLM Backbone=Claude 3.7 Sonnet2025.1163—7886
Mem0LLM Backbone=Claude 3.7 Sonnet2025.1162—7784
Qwen3-Next-4B-InstructN. Params=4B, MoE=false, Reasoning=false2025.1261.6167.2458.72—
WRMDBase Model=Qwen3-Next-4B-Instruct, N. Params=4B, MoE=false, Reasoning=false, Train Dataset=Extended, Top-k Layers=4, Coeff=-0.5, Reposition=false2025.1261.266.1459.32—
LangMemLLM Backbone=Claude 3.7 Sonnet2025.1156—6677
AWMLLM Backbone=Claude 3.7 Sonnet2025.1156—7280
DC-CuLLM Backbone=Claude 3.7 Sonnet2025.1152—6579
DC-RSLLM Backbone=Claude 3.7 Sonnet2025.1152—6079
BaselineLLM Backbone=Gemini 2.5 Flash2025.1146—7583
MemOSLLM Backbone=Gemini 2.5 Flash2025.1146—7582
Mem0LLM Backbone=Gemini 2.5 Flash2025.1146—7483
ReMemLLM Backbone=Gemini 2.5 Flash2025.1146—7985
SelfRAGLLM Backbone=Gemini 2.5 Flash2025.1145—7583
DC-CuLLM Backbone=Gemini 2.5 Flash2025.1144—6979
AWMLLM Backbone=Gemini 2.5 Flash2025.1143—7279
ExpRAGLLM Backbone=Gemini 2.5 Flash2025.1143—7883
HistoryLLM Backbone=Gemini 2.5 Flash2025.1142—7884
AmemLLM Backbone=Gemini 2.5 Flash2025.1142—7283
DC-RSLLM Backbone=Gemini 2.5 Flash2025.1142—6980
LangMemLLM Backbone=Gemini 2.5 Flash2025.1139—7179
ExpRecentLLM Backbone=Gemini 2.5 Flash2025.1139—7583
ReActLLM Backbone=Gemini 2.5 Flash2025.1116—5464