List your company
ResearchFormerly WizWand

Structural QA on StructQA

78AccuracyTAMO
3.608822.921942.23561.5481Nov 30, 2025
Updated 8mo ago

Evaluation results

MethodLinks
TAMOSetting=Frozen LLM, Backbone=Llama 3.1 8B2025.1178
TAMO+Setting=Tuned LLM (SFT)2025.1171.6
Prompt tuningSetting=Frozen LLM, Backbone=Llama 3.1 8B2025.1171.53
TAMO+Setting=Tuned LLM (LoRA)2025.1170.8
SFTSetting=Tuned LLM (SFT)2025.1162.73
GPT-4.1Setting=Others2025.1160.33
TAMOSetting=Frozen LLM2025.1159.07
DeepSeek-R1Setting=Others2025.1157.47
GPT-4Setting=Others2025.1151.4
LoRASetting=Tuned LLM (LoRA)2025.1145.67
GPT-3.5Setting=Others2025.1141.93
Prompt tuningSetting=Frozen LLM2025.1137.8
Llama 3.1 8BSetting=Inference Only2025.1115.73
Zero-shotSetting=Inference Only2025.118.6
TableLlamaSetting=Tuned LLM (SFT)2025.116.47