List your company
ResearchFormerly WizWand

Mathematical Reasoning on MATH (Score)

96.7ScoreQwen 3 VL 32B Think
-3.4540822.5474648.54974.55054Jul 15, 2024Oct 18, 2024Jan 22, 2025Apr 28, 2025Aug 2, 2025Nov 6, 2025Feb 10, 2026
Updated 8mo ago

Evaluation results

MethodLinks
Qwen 3 VL 32B ThinkModel Family=Qwen 3, Parameter Count=32B, Thinking Capability=true2025.1296.7
Olmo 3.1 Think 32BTraining Stage=Final Think 3.1, Model Family=Olmo 3.1, Parameter Count=32B, Thinking Capability=true2025.1296.2
Olmo 3 Think (Final 3.0)Training Stage=Final Think 3.0, Model Family=Olmo 3, Parameter Count=32B, Thinking Capability=true2025.1296.1
Olmo 3 Think (DPO)Training Stage=DPO, Model Family=Olmo 3, Parameter Count=32B, Thinking Capability=true2025.1295.9
Olmo 3 Think (SFT)Training Stage=SFT, Model Family=Olmo 3, Parameter Count=32B, Thinking Capability=true2025.1295.6
Qwen 3 32BModel Family=Qwen 3, Parameter Count=32B, Thinking Capability=false2025.1295.4
Qwen 3 VL 8B Think2025.1295.2
Olmo 3 7B ThinkStage=Final Think2025.1295.1
Qwen 3 8B2025.1295.1
OR Nemotron 7B2025.1294.6
K2-V2 70B InstructModel Family=K2, Parameter Count=70B, Thinking Capability=false2025.1294.5
OpenThinker3 7B2025.1294.5
Olmo 3 7B ThinkStage=SFT2025.1294.4
Nemotron Nano 9B v22025.1294.4
DS-R1 32BModel Family=DeepSeek-R1, Parameter Count=32B, Thinking Capability=true2025.1292.6
Olmo 3 7B ThinkStage=DPO2025.1292.4
DS-R1 Qwen 7B2025.1287.9
GHS-TDABackbone=Llama 3-8B2026.0277.19
AoTBackbone=Llama 3-8B2026.0276.91
GoTBackbone=Llama 3-8B2026.0276.36
ToTBackbone=Llama 3-8B2026.0272.86
CoTBackbone=Llama 3-8B2026.0272.04
GHS-TDABackbone=Qwen 2-14B2026.0268.4
AoTBackbone=Qwen 2-14B2026.0266.2
GoTBackbone=Qwen 2-14B2026.0265.7
ToTBackbone=Qwen 2-14B2026.0265
CoTBackbone=Qwen 2-14B2026.0260.8
LLM-in-Sandbox-RLBackbone=Qwen3-4B-Instruct-2507, Sandbox Mode=true2026.0153.3
LLM-in-Sandbox-RLBackbone=Qwen3-4B-Instruct-2507, Sandbox Mode=false2026.0151
LLM-RLBackbone=Qwen3-4B-Instruct-2507, Sandbox Mode=false2026.0148.5
Base LLMBackbone=Qwen3-4B-Instruct-2507, Sandbox Mode=false2026.0146
Qwen2-7BParameters=7B2024.0744.2
LLM-in-Sandbox-RLBackbone=Qwen3-Coder-30B-A3B, Sandbox Mode=true2026.0142.7
Base LLMBackbone=Qwen3-Coder-30B-A3B, Sandbox Mode=true2026.0141.5
LLM-RLBackbone=Qwen3-Coder-30B-A3B, Sandbox Mode=true2026.0140.9
LLM-RLBackbone=Qwen3-4B-Instruct-2507, Sandbox Mode=true2026.0133.1
Base LLMBackbone=Qwen3-4B-Instruct-2507, Sandbox Mode=true2026.0132.5
LLM-in-Sandbox-RLBackbone=Qwen3-Coder-30B-A3B, Sandbox Mode=false2026.0129.8
LLM-RLBackbone=Qwen3-Coder-30B-A3B, Sandbox Mode=false2026.0128.1
Base LLMBackbone=Qwen3-Coder-30B-A3B, Sandbox Mode=false2026.0126
Gemma-7BParameters=7B2024.0724.3
Llama-3-8BParameters=8B2024.0720.5
Qwen1.5-7BParameters=7B2024.0720.3
Mistral-7BParameters=7B2024.0713.1
Phi-42026.010.801
Llama-3.3-70B-InstructParameters=70B2026.010.739
Llama-3.1-8B-InstructParameters=8B2026.010.479
Foundation-Sec-8B-InstructParameters=8B2026.010.436
Foundation-Sec-8B-ReasoningParameters=8B2026.010.433
Llama-Primus-Merged2026.010.398