List your company

Mathematical Calculation on AQuA-RAT

83.1Accuracy (AQuA-RAT)MemAPO
58.76465.08271.477.718Mar 23, 2026

Evaluation results

MethodLinks
MemAPOBackbone=Qwen3-8B2026.0383.1
CoTBackbone=Qwen3-8B2026.0382.5
OPROBackbone=Qwen3-8B2026.0382.5
MemAPOBackbone=GPT-4o-mini2026.0382.5
Step-BackBackbone=Qwen3-8B2026.0381.8
ProTeGiBackbone=Qwen3-8B2026.0381.2
RaRBackbone=Qwen3-8B2026.0380.5
TextGradBackbone=Qwen3-8B2026.0380.5
IOBackbone=Qwen3-8B2026.0379.9
TextGradBackbone=GPT-4o-mini2026.0379.9
OPROBackbone=GPT-4o-mini2026.0379.2
PromptBreederBackbone=GPT-4o-mini2026.0378.6
SPOBackbone=GPT-4o-mini2026.0377.9
ProTeGiBackbone=GPT-4o-mini2026.0376
PromptBreederBackbone=Qwen3-8B2026.0374
SPOBackbone=Qwen3-8B2026.0372.7
CoTBackbone=GPT-4o-mini2026.0369.5
RaRBackbone=GPT-4o-mini2026.0366.9
IOBackbone=GPT-4o-mini2026.0361.7
Step-BackBackbone=GPT-4o-mini2026.0359.7