List your company
ResearchFormerly WizWand

Reasoning on DeepScaler

57.3AccuracyUAB
14.6625.7336.847.87May 26, 2026
Updated 5mo ago

Evaluation results

MethodLinks
UABBackbone Model=Cohere, Budget=N=42026.0557.3
LLM-JudgeBackbone Model=Cohere, Budget=N=42026.0557.1
UniformBackbone Model=Cohere, Budget=N=42026.0556.8
N=1Backbone Model=Cohere, Budget=N=12026.0553.7
UABBackbone Model=Gemma3-27B, Budget=N=42026.0541.1
LLM-JudgeBackbone Model=Gemma3-27B, Budget=N=42026.0540.3
UniformBackbone Model=Gemma3-27B, Budget=N=42026.0540.2
UABModel=Qwen2.5-7B, N=42026.0538.1
UABBackbone Model=GPT-OSS-20B, Budget=N=42026.0536.4
RandomModel=Qwen2.5-7B, N=42026.0536.1
LLM-JudgeBackbone Model=GPT-OSS-20B, Budget=N=42026.0535.1
UniformBackbone Model=GPT-OSS-20B, Budget=N=42026.0534.7
LLM-JudgeModel=Qwen2.5-7B, N=42026.0534.3
N=1Backbone Model=Gemma3-27B, Budget=N=12026.0533.7
UniformModel=Qwen2.5-7B, N=42026.0533.3
LengthModel=Qwen2.5-7B, N=42026.0527.8
N=1Model=Qwen2.5-7B, N=12026.0527.5
N=1Backbone Model=GPT-OSS-20B, Budget=N=12026.0525.5
UABModel=Qwen2.5-1.5B, N=42026.0522.4
RandomModel=Qwen2.5-1.5B, N=42026.0521.4
UABModel=Llama3.2-3B, N=42026.0521.4
LengthModel=Qwen2.5-1.5B, N=42026.0521.2
LengthModel=Llama3.2-3B, N=42026.0520.2
UniformModel=Qwen2.5-1.5B, N=42026.0519.2
LLM-JudgeModel=Qwen2.5-1.5B, N=42026.0518.8
RandomModel=Llama3.2-3B, N=42026.0518.8
UniformModel=Llama3.2-3B, N=42026.0517.9
N=1Model=Qwen2.5-1.5B, N=12026.0517.4
N=1Model=Llama3.2-3B, N=12026.0516.3
LLM-JudgeModel=Llama3.2-3B, N=42026.0516.3