List your company

Multi-Turn Tool-Integrated Reasoning (TIR) on AIME25 (Peak avg@32 score)

28.13Peak avg@32 ScoreOTB
18.156420.745723.33525.9243Feb 6, 2026

Evaluation results

MethodLinks
OTBBackbone=Qwen2.5-7B, Reasoning Paradigm=TIR2026.0228.13
SimpleTIRBackbone=Qwen2.5-7B, Reasoning Paradigm=TIR2026.0226.67
OGBBackbone=Qwen2.5-7B, Reasoning Paradigm=TIR2026.0221.15
OPOBackbone=Qwen2.5-7B, Reasoning Paradigm=TIR2026.0220.83
RLOOBackbone=Qwen2.5-7B, Reasoning Paradigm=TIR2026.0220.1
GRPOBackbone=Qwen2.5-7B, Reasoning Paradigm=TIR2026.0218.54