List your company

Agent Tool Use on T-eval (Held-Out)

71.8AccuracySHAD+RFT
53.658.32563.0567.775Dec 19, 2024

Evaluation results

MethodLinks
SHAD+RFTModel=LLaMA3-8B2024.1271.8
SHAD+α-FTModel=LLaMA3-8B2024.1268.8
Rho-1Model=LLaMA3-8B2024.1268.4
SHAD+RFTModel=LLaMA3.1-8B2024.1268.3
SFTModel=LLaMA3-8B2024.1267
Rho-1Model=LLaMA3.1-8B2024.1267
RewardFTModel=LLaMA3.1-8B2024.1266.4
RewardFTModel=LLaMA3-8B2024.1266.3
SFTModel=LLaMA3.1-8B2024.1264.2
SHAD+α-FTModel=LLaMA3.1-8B2024.1263.8
Regex+RFTModel=LLaMA3-8B2024.1261.1
RegexModel=LLaMA3.1-8B2024.1258.6
Regex+RFTModel=LLaMA3.1-8B2024.1257.6
RegexModel=LLaMA3-8B2024.1254.3