List your company

Reasoning on AIME 25

76.9AccuracyParallel-Probe
18.86833.9344964.066Nov 13, 2025Nov 26, 2025Dec 10, 2025Dec 24, 2025Jan 6, 2026Jan 20, 2026Feb 3, 2026

Evaluation results

MethodLinks
Parallel-ProbeBase Model=Qwen3-8B, Type=Parallel2026.0276.921.9846.7
SC@64Base Model=Qwen3-8B, Type=Parallel2026.0276.730.71,124.4
ASCBase Model=Qwen3-8B, Type=Seq.2026.0276.7406.2406.2
ESCBase Model=Qwen3-8B, Type=Hybrid2026.0276.7132.4793.1
SC@64Base Model=Qwen3-4B, Type=Parallel2026.0276.630.51,088.1
ASCBase Model=Qwen3-4B, Type=Seq.2026.0276.6325.1325.1
ESCBase Model=Qwen3-4B, Type=Hybrid2026.0276.6137793.3
Parallel-ProbeBase Model=Qwen3-4B, Type=Parallel2026.0276.122.2806
SC@64 + SACBase Model=Qwen3-4B, Type=Parallel2026.0273.327.9995.4
FP16Backbone=Qwen3-8B, Zero-shot=true2025.1172.2——
SC@64 + SACBase Model=Qwen3-8B, Type=Parallel2026.0270.228.1998.5
QTIPBackbone=Qwen3-14B, Type=vector, Zero-shot=true2025.1169——
FP16Backbone=Qwen3-14B, Zero-shot=true2025.1168.9——
AWQBackbone=Qwen3-14B, Type=linear, Zero-shot=true2025.1168.9——
ParoQuantBackbone=Qwen3-14B, Type=linear, Zero-shot=true2025.1167.8——
ParoQuantBackbone=Qwen3-4B, Type=linear, Zero-shot=true2025.1163.3——
QTIPBackbone=Qwen3-8B, Type=vector, Zero-shot=true2025.1163.3——
ParoQuantBackbone=Qwen3-8B, Type=linear, Zero-shot=true2025.1163.3——
FP16Backbone=Qwen3-4B, Zero-shot=true2025.1162.2——
AWQBackbone=Qwen3-8B, Type=linear, Zero-shot=true2025.1161.1——
E-QATBackbone=Qwen3-14B, Type=linear, Zero-shot=true2025.1161.1——
QTIPBackbone=Qwen3-4B, Type=vector, Zero-shot=true2025.1158.9——
AWQBackbone=Qwen3-4B, Type=linear, Zero-shot=true2025.1153.3——
E-QATBackbone=Qwen3-8B, Type=linear, Zero-shot=true2025.1151.1——
E-QATBackbone=Qwen3-4B, Type=linear, Zero-shot=true2025.1146.7——
Parallel-ProbeBase Model=Qwen3-1.7B, Type=Parallel2026.0244.721.3775.8
SC@64Base Model=Qwen3-1.7B, Type=Parallel2026.0244.4301,054.1
ASCBase Model=Qwen3-1.7B, Type=Seq.2026.0244.4600.9600.9
ESCBase Model=Qwen3-1.7B, Type=Hybrid2026.0244.4160.6913.8
SC@64 + SACBase Model=Qwen3-1.7B, Type=Parallel2026.024026.4909
FP16Backbone=R1-Distill-Llama-8B, Zero-shot=true2025.1132.2——
ParoQuantBackbone=R1-Distill-Llama-8B, Type=linear, Zero-shot=true2025.1131.1——
QTIPBackbone=R1-Distill-Llama-8B, Type=vector, Zero-shot=true2025.1130.1——
Parallel-ProbeBase Model=Qwen3-0.6B, Type=Parallel2026.0229.719.6697.8
SC@64Base Model=Qwen3-0.6B, Type=Parallel2026.0228.931.1890.5
ASCBase Model=Qwen3-0.6B, Type=Seq.2026.0228.9653.8653.8
ESCBase Model=Qwen3-0.6B, Type=Hybrid2026.0228.9171.8868.8
AWQBackbone=R1-Distill-Llama-8B, Type=linear, Zero-shot=true2025.1126.7——
SC@64 + SACBase Model=Qwen3-0.6B, Type=Parallel2026.0225.427.2819.4
E-QATBackbone=R1-Distill-Llama-8B, Type=linear, Zero-shot=true2025.1121.1——