INDEX

Explanations

safety and efficacy

The neuron is tuned to spotting clinical‐trial outcome terms such as “safety,” “efficacy,” and “tolerability.”

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

-1.59

。」

-1.41

PARATION

-1.39

＇

-1.39

赑

-1.30

ݮ

-1.28

’’

-1.28

It

-1.27

 laude

-1.27

springframework

-1.26

POSITIVE LOGITS

and

1.95

1.45

éle

1.38

 小学生

1.34

muñeca

1.29

 Nachricht

1.28

kasi

1.27

 أنا

1.26

 proberen

1.26

 või

1.24

Activations Density 0.043%