List your company

Question Answering on TriviaQA In-Distribution

96.09Exact Match (EM)SkillFlow
42.831656.658370.48584.3117May 13, 2026

Evaluation results

MethodLinks
SkillFlowMethod Category=Ours2026.0596.0996.94
Qwen3.5-AFlowMethod Category=AFlow, Backbone=Qwen3.52026.0592.9793.25
FlowSteerMethod Category=Agent+RL2026.0591.4192.06
AgentFlowMethod Category=Agent+RL2026.0589.8490.47
SkillRLMethod Category=Agent+RL2026.0589.0690.63
v4-flashMethod Category=Baseline2026.0574.0281.99
Qwen3.5-SFTMethod Category=SFT, Backbone=Qwen3.52026.0547.2455.73
Qwen3.5-GRPOMethod Category=GRPO, Backbone=Qwen3.52026.0547.2456.93
Qwen3.5-9BMethod Category=Baseline2026.0544.8854.3