List your company

Question Answering on OfficeQA 246 questions

80.1AccuracySkillSmith
9.06827.50945.9564.391May 31, 2026

Evaluation results

MethodLinks
SkillSmithBackbone=Qwen3.5-397B2026.0580.1197
EvoSkillBackbone=Qwen3.5-397B2026.0571.5176
Base AgentBackbone=Qwen3.5-397B2026.0561.8152
SkillSmithBackbone=Qwen3.5-122B2026.0560.2148
EvoSkillBackbone=Qwen3.5-122B2026.0553.3131
Base AgentBackbone=Qwen3.5-122B2026.0543.9108
SkillSmithBackbone=Qwen3.5-35B2026.0538.695
EvoSkillBackbone=Qwen3.5-35B2026.0534.184
Base AgentBackbone=Qwen3.5-35B2026.0528.971
SkillSmithBackbone=Qwen3.5-27B2026.0528.570
EvoSkillBackbone=Qwen3.5-27B2026.0525.663
Base AgentBackbone=Qwen3.5-27B2026.052254
SkillSmithBackbone=Qwen3.5-9B2026.0514.636
EvoSkillBackbone=Qwen3.5-9B2026.0513.433
Base AgentBackbone=Qwen3.5-9B2026.0511.829