List your company

Reasoning on GPQA Ambiguity-Augmented (subset of 200 samples)

44Accuracy@1DisambiguSLM
38.69640.07341.4542.827Apr 25, 2026

Evaluation results

MethodLinks
DisambiguSLM2026.044447.311.2
SPO2026.0443.646.817.5
OPRO2026.0443.346.518.9
Step-back2026.0442.44520.8
CoT2026.0441.643.922.1
PromptAgent2026.0441.344.120.4
APE2026.0441.143.821.6
PromptBreeder2026.0440.94321.1
Rephrase2026.0440.242.123.4
TextGrad2026.0440.242.622.3
Naïve prompting2026.0438.940.224.7