HyperAI
Command Palette
Search for a command to run...
このベンチマークにおける各種モデルの性能結果
メトリクス
Tool Use
MultiModal
Trustworthiness
w/o Fine Tuning
Long-Term Memory
Data Understanding
Numerical Analysis
Knowledge& Reasoning
Tool-wise Reflection
MM.
Num.
Data.
Know.
Total
Trust.
Avg.
Methods
Response
RAD
MMIU
eICU
MedQA
SLAKE
emrQA
BioASQ
EHR SQL
MedCalc
HealthQA
EHR -Halt
Long Halt
Pub MedQA
Avg. (Num)
MIMIC -III
Omni MedQA
Total Avg.
Avg. (Trust)
Med Mentions
MedHalt -Rht
MedVQA -Halt
Avg. (Knowledge)
Long Avg. (Data)
8 行 合計