List your company

Reasoning on BBH (pass@1)

83.69BBH Pass@1Learned
1.873223.114144.35565.5959Jun 12, 2025Aug 8, 2025Oct 5, 2025Dec 2, 2025Jan 29, 2026Mar 28, 2026May 25, 2026

Evaluation results

MethodLinks
LearnedBackbone Model=DS-R1-Llama-8B, Cmax=3, Bmax=92162025.0683.69
WaitBackbone Model=DS-R1-Llama-8B, Cmax=3, Bmax=92162025.0681.97
HmmBackbone Model=DS-R1-Llama-8B, Cmax=3, Bmax=92162025.0681.79
AlternativelyBackbone Model=DS-R1-Llama-8B, Cmax=3, Bmax=92162025.0681.32
CritiqueBackbone Model=DS-R1-Llama-8B, Cmax=3, Bmax=92162025.0680.26
Baseline (w/o BF)Backbone Model=DS-R1-Llama-8B, Cmax=3, Bmax=92162025.0679.73
LearnedBackbone Model=DS-R1-Qwen-7B, Cmax=3, Bmax=92162025.0674.75
WaitBackbone Model=DS-R1-Qwen-7B, Cmax=3, Bmax=92162025.0673.85
HmmBackbone Model=DS-R1-Qwen-7B, Cmax=3, Bmax=92162025.0672.91
AlternativelyBackbone Model=DS-R1-Qwen-7B, Cmax=3, Bmax=92162025.0672.53
Baseline (w/o BF)Backbone Model=DS-R1-Qwen-7B, Cmax=3, Bmax=92162025.0671.05
CritiqueBackbone Model=DS-R1-Qwen-7B, Cmax=3, Bmax=92162025.0671.01
SCF-RKLModels=Fuse2026.0269.92
w/o MergingModels=Code2026.0269.29
Dare Task ArithmeticModels=Fuse2026.0266.17
Task ArithmeticModels=Fuse2026.0266.15
w/o MergingModel=Meta-Llama-3-8B-Instruct (base0)2026.0265.78
Dare Ties MergingModels=Fuse2026.0264.75
Dare Ties MergingModel=Fused2026.0263.82
Task ArithmeticModel=Fused2026.0263.63
Ties MergingModels=Fuse2026.0260.7
SCEModels=Fuse2026.0260.53
Ties MergingModel=Fused2026.0257.89
w/o MergingModel=MAmmoTH2-8B-Plus (base1)2026.0253.92
TeacherModel=Gemma3-4B2026.0552.57
SCF-RKLModel=Fused2026.0251.64
w/o MergingModels=Math2026.0248.15
LearnedBackbone Model=DS-R1-Qwen-1.5B, Cmax=3, Bmax=92162025.0647.7
WaitBackbone Model=DS-R1-Qwen-1.5B, Cmax=3, Bmax=92162025.0647.58
HmmBackbone Model=DS-R1-Qwen-1.5B, Cmax=3, Bmax=92162025.0647.06
AlternativelyBackbone Model=DS-R1-Qwen-1.5B, Cmax=3, Bmax=92162025.0646.75
CritiqueBackbone Model=DS-R1-Qwen-1.5B, Cmax=3, Bmax=92162025.0646
Baseline (w/o BF)Backbone Model=DS-R1-Qwen-1.5B, Cmax=3, Bmax=92162025.0645.71
Dare Task ArithmeticModel=Fused2026.0226.98
SCEModel=Fused2026.0224.48
Reverse KLDistillation Family=KL Divergence Family2026.059.52
Bridge-Garden (Skew FKL)Distillation Family=Distance & Skewed Family2026.057.78
Bridge-Garden (α-β divergence)Distillation Family=Unified Divergence Family2026.057.56
α-β divergenceDistillation Family=Unified Divergence Family2026.057.26
Bridge-Garden (Reverse KL)Distillation Family=KL Divergence Family2026.057.14
Skew FKLDistillation Family=Distance & Skewed Family2026.057.01
Student (No Distill)Model=Gemma3-1B2026.056.47
Total VariationDistillation Family=Distance & Skewed Family2026.056.29
Bridge-Garden (Forward KL)Distillation Family=KL Divergence Family2026.056.23
Forward KLDistillation Family=KL Divergence Family2026.055.91
Skew RKLDistillation Family=Distance & Skewed Family2026.055.83
Bridge-Garden (Skew RKL)Distillation Family=Distance & Skewed Family2026.055.72
Hard KDTeacher Model=Gemma3-4B, Student Model=Gemma3-1B2026.055.28
Bridge-Garden (Total Variation)Distillation Family=Distance & Skewed Family2026.055.02