List your company

Multi-Hop Question Answering on Bamboogle (F1, Collapse Point)

62.25F1 ScoreCalibAdv
45.277249.683654.0958.4964Apr 20, 2026

Evaluation results

MethodLinks
CalibAdvBackbone=Qwen2.5-7B-Base2026.0462.25—
LLDBackbone=Qwen2.5-7B-Base2026.0458.36—
StepSearchBackbone=Qwen2.5-7B-Base2026.0455.74—
CalibAdvBackbone=Llama3.2-3B-Instruct2026.0454.2—
LLDBackbone=Llama3.2-3B-Instruct2026.0453.34—
MT-GRPOBackbone=Qwen2.5-7B-Base2026.0451.47—
SimpleTIRBackbone=Qwen2.5-7B-Base2026.0450.965
CalibAdvBackbone=Qwen2.5-3B-Base2026.0450.62—
Search-R1Backbone=Qwen2.5-3B-Base2026.0450.48292
Search-R1Backbone=Qwen2.5-7B-Base2026.0448.6658
SimpleTIRBackbone=Qwen2.5-3B-Base2026.0448.58310
Search-R1Backbone=Llama3.2-3B-Instruct2026.0447.3340
GiGPOBackbone=Qwen2.5-7B-Base2026.0447.11—
SimpleTIRBackbone=Llama3.2-3B-Instruct2026.0445.9354