List your company

Grade School Math Reasoning on GSM8K (Accuracy, Latency)

94.53AccuracyAgentDropout
63.444471.514779.58587.6553May 28, 2026

Evaluation results

MethodLinks
AgentDropoutBackbone=Qwen2.5-14B-Instruct, Topology=Star, Training-Free (TF)=false2026.0594.5388.62
LLM-DebateBackbone=Qwen2.5-14B-Instruct, Topology=Debate, Training-Free (TF)=true2026.0594.24112.34
AgentDropoutBackbone=Qwen2.5-14B-Instruct, Topology=Layered, Training-Free (TF)=false2026.0594.2495.85
Vanilla MASBackbone=Qwen2.5-14B-Instruct, Topology=Layered, Training-Free (TF)=true2026.0594.14119.64
AgentDropoutBackbone=Qwen2.5-14B-Instruct, Topology=Random, Training-Free (TF)=false2026.0594.1485.58
CONCATBackbone=Qwen2.5-14B-Instruct, Topology=Hybrid, Training-Free (TF)=true2026.0594.151.19
Vanilla MASBackbone=Qwen2.5-14B-Instruct, Topology=Star, Training-Free (TF)=true2026.0594.06131.6
SC-CoTBackbone=Qwen2.5-72B-Instruct, Topology=-, Training-Free (TF)=true2026.0594.06131.48
AgentDropoutBackbone=Qwen2.5-14B-Instruct, Topology=Chain, Training-Free (TF)=false2026.0594.0182.01
Vanilla MASBackbone=Qwen2.5-14B-Instruct, Topology=Random, Training-Free (TF)=true2026.0593.98117.23
Vanilla MASBackbone=Qwen2.5-14B-Instruct, Topology=FullConnected, Training-Free (TF)=true2026.0593.98132.1
AgentDropoutBackbone=Qwen2.5-14B-Instruct, Topology=FullConnected, Training-Free (TF)=false2026.0593.9893.51
LLM-DebateBackbone=Qwen2.5-72B-Instruct, Topology=Debate, Training-Free (TF)=true2026.0593.93303.27
CoTBackbone=Qwen2.5-72B-Instruct, Topology=-, Training-Free (TF)=true2026.0593.8537.26
AgentDropoutBackbone=Qwen2.5-72B-Instruct, Topology=FullConnected, Training-Free (TF)=false2026.0593.85244.47
Vanilla MASBackbone=Qwen2.5-72B-Instruct, Topology=Layered, Training-Free (TF)=true2026.0593.78312.79
AgentDropoutBackbone=Qwen2.5-72B-Instruct, Topology=Star, Training-Free (TF)=false2026.0593.72232.63
Vanilla MASBackbone=Qwen2.5-14B-Instruct, Topology=Chain, Training-Free (TF)=true2026.0593.57110.68
Vanilla MASBackbone=Qwen2.5-72B-Instruct, Topology=Star, Training-Free (TF)=true2026.0593.49341.8
CONCATBackbone=Qwen2.5-72B-Instruct, Topology=Hybrid, Training-Free (TF)=true2026.0593.49141.72
Vanilla MASBackbone=Qwen2.5-72B-Instruct, Topology=Random, Training-Free (TF)=true2026.0593.44308.52
AgentDropoutBackbone=Qwen2.5-72B-Instruct, Topology=Chain, Training-Free (TF)=false2026.0593.36228.89
Vanilla MASBackbone=Qwen2.5-72B-Instruct, Topology=FullConnected, Training-Free (TF)=true2026.0593.28334.57
AgentDropoutBackbone=Qwen2.5-72B-Instruct, Topology=Layered, Training-Free (TF)=false2026.0593.26243.1
SC-CoTBackbone=Qwen2.5-14B-Instruct, Topology=-, Training-Free (TF)=true2026.0593.0554
Vanilla MASBackbone=Qwen2.5-72B-Instruct, Topology=Chain, Training-Free (TF)=true2026.0592.86299.55
AgentDropoutBackbone=Qwen2.5-72B-Instruct, Topology=Random, Training-Free (TF)=false2026.0592.76231.72
LLM-DebateBackbone=Llama3-8B-Instruct, Topology=Debate, Training-Free (TF)=true2026.0578.8133.92
AgentDropoutBackbone=Llama3-8B-Instruct, Topology=Chain, Training-Free (TF)=false2026.0578.5785.37
CONCATBackbone=Llama3-8B-Instruct, Topology=Hybrid, Training-Free (TF)=true2026.0577.0872.02
Vanilla MASBackbone=Llama3-8B-Instruct, Topology=Layered, Training-Free (TF)=true2026.0576.35158.44
AgentDropoutBackbone=Llama3-8B-Instruct, Topology=Random, Training-Free (TF)=false2026.0576.3298.48
AgentDropoutBackbone=Llama3-8B-Instruct, Topology=Star, Training-Free (TF)=false2026.0576.28107.69
AgentDropoutBackbone=Llama3-8B-Instruct, Topology=Layered, Training-Free (TF)=false2026.0576.25100.57
Vanilla MASBackbone=Llama3-8B-Instruct, Topology=FullConnected, Training-Free (TF)=true2026.0575.6149.22
Vanilla MASBackbone=Llama3-8B-Instruct, Topology=Random, Training-Free (TF)=true2026.0575.57135.14
Vanilla MASBackbone=Llama3-8B-Instruct, Topology=Star, Training-Free (TF)=true2026.0575.49147.11
AgentDropoutBackbone=Llama3-8B-Instruct, Topology=FullConnected, Training-Free (TF)=false2026.0575.09118.11
SC-CoTBackbone=Llama3-8B-Instruct, Topology=-, Training-Free (TF)=true2026.0574.1940.77
Vanilla MASBackbone=Llama3-8B-Instruct, Topology=Chain, Training-Free (TF)=true2026.0573.7122.11
CoTBackbone=Qwen2.5-14B-Instruct, Topology=-, Training-Free (TF)=true2026.0567.9713.78
CoTBackbone=Llama3-8B-Instruct, Topology=-, Training-Free (TF)=true2026.0564.649.2