STRING on EHRCon benchmark leaderboard · SOTA2 Research

STRING on EHRCon

67.4LLM Judge AccuracyA3

Evaluation Results

Method	Links
A32026.05		67.4
A32026.05		66.7
ReACT2026.05		61.5
GiGPO2026.05		59.3
RetroAgent2026.05		57.2
HGPO2026.05		54.1
GSPO2026.05		45.6
LATS2026.05		41.5
rStar2026.05		36.7