Tool Selection on MetaTool 199 tools, 1,287 queries (30% test) benchmark leaderboard · SOTA2 Research

Tool Selection on MetaTool 199 tools, 1,287 queries (30% test)

83R@1OATS-S1

Evaluation Results

Method	Links
OATS-S12026.03		83	96.4	98.6	94
OATS-S32026.03		81	95.6	98.8	93.1
SE2026.03		71.6	90.4	95.8	86.9
OATS-S22026.03		71.6	90.9	95.4	86.9
SE + Lexical2026.03		64	86	92.7	81.6
BM252026.03		39.7	61.7	75.3	59.5
Random2026.03		9.6	30.8	49.2	29.8