List your company

Question Answering on NQ, TriviaQA, PopQA, HotpotQA, 2wiki, MuSiQue, Bamboogle

48.2Average QA ScoreIGRPO
-1.262411.578824.4237.2612Aug 27, 2025Oct 18, 2025Dec 9, 2025Jan 31, 2026Mar 24, 2026May 15, 2026Jul 7, 2026

Evaluation results

MethodLinks
IGRPOBackbone=Qwen2.5-7B-Instruct, Type=Tree-based2026.0748.2——————————
SAPOBackbone=Qwen2.5-7B-Instruct2026.0647.848.462.946.74545.218.346.4———
GiGPOBackbone=Qwen2.5-7B-Instruct, Type=Chain-based2026.0747.3——————————
IGPOBackbone=Qwen2.5-7B-Instruct, Type=Chain-based2026.0745.7——————————
SKILLRLBackbone=Qwen2.5-7B-Instruct2026.0645.545.961.245.943.939.917.245.6———
IGRPOBackbone=Qwen2.5-3B-Instruct, Type=Tree-based2026.0745.4——————————
ZeroSearchBackbone=Qwen2.5-7B-Instruct, Type=Chain-based2026.0744.8——————————
Skill0Backbone=Qwen2.5-7B-Instruct2026.0644.442.763.645.34038.316.466.9———
Teacher-8BModel Family=Llama 3, Configuration=Teacher model2025.0843.847.564.744.842.744.317.944.4———
Tree-GRPOBackbone=Qwen2.5-7B-Instruct, Type=Tree-based2026.0743.6——————————
EvolveRBackbone=Qwen2.5-7B-Instruct2026.0643.143.563.444.638.24215.654.4———
Search-R1Backbone=Qwen2.5-7B-Instruct, Type=Chain-based2026.0742.5——————————
GiGPOBackbone=Qwen2.5-3B-Instruct, Type=Chain-based2026.0742.3——————————
IGPOBackbone=Qwen2.5-3B-Instruct, Type=Chain-based2026.0741.4——————————
SLIM2026.054138.662.24037.231.712.837.637.936.9—
SLIM†variant=†2026.054138.462.140.436.931.512.73637.736.5—
GRPOTeacher model=Qwen3-8B, Role=Teacher baseline2026.074136.659.340.236.734.713.839.2———
Tree-GRPOBackbone=Qwen2.5-3B-Instruct, Type=Tree-based2026.0740.8——————————
OPDTeacher model=Qwen3-4B-Instruct-2507, Student model=Qwen3-4B-Instruct-25072026.0740.636.359.740.337.331.714.544———
ReOPDTeacher model=Qwen3-4B-Instruct-2507, Student model=Qwen3-4B-Instruct-25072026.0740.536.459.640.537.331.51437.6———
GRPOTeacher model=Qwen3-4B-Instruct-2507, Role=Teacher baseline2026.0740.436.159.240.237.231.813.839.2———
GRPO-SGBackbone=Qwen3-4B, Search Framework=VT-Search2025.1040.1848.2364.114636.630.8910.2641.6——14.1
Skill02026.0539.337.959.538.632.731.910.332.835.334.6—
ZeroSearchBackbone=Qwen2.5-7B-Instruct2026.0639.143.661.851.534.635.218.427.8———
OPDTeacher model=Qwen3-8B, Student model=Qwen3-4B-Instruct-25072026.0739.135.557.3393531.212.837.6———
ZeroSearchBackbone=Qwen2.5-3B-Instruct, Type=Chain-based2026.0739.1——————————
ReOPDTeacher model=Qwen3-8B, Student model=Qwen3-4B-Instruct-25072026.073934.957.539.134.63111.937.6———
DGPOModel Family=Llama 3, Distillation Configuration=8B → 1B2025.0838.944.855.343.741.237.915.533.9———
Search-R1Backbone=Qwen2.5-7B-Instruct2026.0638.539.36139.73740.114.636.8———
SkillRL†variant=†2026.0538.136.859.836.931.529.710.331.234.233.6—
GRPO2026.0537.535.957.836.530.830.19.230.433.432.8—
Search-R1Backbone=Qwen2.5-3B-Instruct, Type=Chain-based2026.0737.4——————————
ARBackbone=Qwen3-4B, Search Framework=VT-Search2025.1036.1149.5561.4536.9838.1923.958.2534.4——2.6
BaseTeacher model=Qwen3-4B-Instruct-2507, Student model=Qwen3-4B-Instruct-25072026.0735.628.455.530.630.833.68.435.8———
80/20Backbone=Qwen3-4B, Search Framework=VT-Search2025.1035.447.9462.6740.2737.5225.666.5127.2——0.5
GRPOBackbone=Qwen3-4B, Search Framework=VT-Search2025.1035.2146.4859.8440.3935.5428.8710.2627.2———
KDModel Family=Llama 3, Distillation Configuration=8B → 1B2025.0834.740.650.840.536.935.511.926.6———
LoptiBackbone=Qwen3-4B, Search Framework=VT-Search2025.1034.6246.2162.0939.1136.7425.9412.2420——1.7
Cold StartTeacher model=Qwen3-4B-Instruct-2507, Student model=Qwen3-4B-Instruct-25072026.0732.126.248.633.624.726.65.824.8———
R1-InstructBackbone=Qwen2.5-3B-Instruct, Type=RL Training2026.0729.8——————————
Qwen3-4BBackbone=Qwen3-4B, Search Framework=VT-Search2025.1028.9531.686038.627.5914.896.723.2———
GRPO-SGBackbone=Qwen2.5-3B, Search Framework=VT-Search2025.1027.2941.852.339.124.520.358——97.2
R1-InstructBackbone=Qwen2.5-7B-Instruct, Type=RL Training2026.0725.8——————————
RAGBackbone=Qwen2.5-7B-Instruct2026.0625.527.458.217.825.823.29.416.8———
PPOModel Family=Llama 3, Distillation Configuration=8B → 1B2025.082535.449.939.422.218.13.76.5———
ARBackbone=Qwen2.5-3B, Search Framework=VT-Search2025.1023.2834.1744.5130.2220.1522.184.547.2——68.2
R1-InstructBackbone=Qwen2.5-7B-Instruct2026.0622.42144.917.120.827.5619.2———
Search-o1Backbone=Qwen2.5-7B-Instruct2026.0622.119.440.611.417278.630.4———
LoptiBackbone=Qwen2.5-3B, Search Framework=VT-Search2025.1021.6240.2240.1925.616.5318.224.186.4——56.2
80/20Backbone=Qwen2.5-3B, Search Framework=VT-Search2025.1018.625.9538.5517.2818.5414.33.6112——34.4
CoTBackbone=Qwen2.5-7B-Instruct2026.0617.412.835.63.816.222.66.624———
Qwen2.5Backbone=Qwen2.5-7B-Instruct2026.0615.211.635.61.216.422.24.814.4———
GRPOBackbone=Qwen2.5-3B, Search Framework=VT-Search2025.1013.8413.529.411.514.516.31.310.4———
Student-1BModel Family=Llama 3, Configuration=Student baseline2025.083.95.284.42.74.20.12.4———
Qwen2.5-3BBackbone=Qwen2.5-3B, Search Framework=VT-Search2025.100.640.51.20.80.51.500———
StepSearchBackbone=Qwen2.5-7B-Instruct2026.06————38.636.622.640———