List your company

Knowledge-intensive Reasoning on GPQA ambiguity-augmented

42.8AccuracyDisambiguSLM
36.24837.94939.6541.351Apr 25, 2026

Evaluation results

MethodLinks
DisambiguSLM2026.0442.8
SPO2026.0441.1
OPRO2026.0440.3
Step-back2026.0439.7
PromptAgent2026.0439.2
CoT2026.0438.9
TextGrad2026.0438.9
APE2026.0438.8
PromptBreeder2026.0438.5
Rephrase2026.0437.1
Naïve prompting2026.0436.5