List your company

Question Answering on DROP (Acc, F1, Avg. Tokens (K))

81.08AccuracySmolagent
26.604840.747454.8969.0326Oct 30, 2025Dec 4, 2025Jan 9, 2026Feb 14, 2026Mar 21, 2026Apr 26, 2026Jun 1, 2026

Evaluation results

MethodLinks
Smolagent2025.1081.0881.0812.01
OWL2025.1079.8579.8511.47
SUPERVISORAGENTmodule=SMAS, base_agent=Smolagent2025.1079.879.811.34
MetaAgent2025.1078.1678.161.43
CoT SCshots=3-shot2025.1077.7277.722.73
Vanilla2025.1076.3676.360.46
GRZOModel=Llama3-8B, Optimization Type=ZO2026.0665——
AdamModel=Llama3-8B, Optimization Type=FO2026.0659.4——
LoRAModel=Llama3-8B, Optimization Type=FO2026.0658.2——
FZOOModel=Llama3-8B, Optimization Type=ZO2026.0657.4——
MeZOModel=Llama3-8B, Optimization Type=ZO2026.0657.1——
GRZOModel=OPT-13B, Optimization Type=ZO2026.0632.8——
AdamModel=OPT-13B, Optimization Type=FO2026.0631.3——
LoRAModel=OPT-13B, Optimization Type=FO2026.0630.9——
MeZOModel=OPT-13B, Optimization Type=ZO2026.0630.9——
FZOOModel=OPT-13B, Optimization Type=ZO2026.0628.7——