INDEX

Explanations

effect of modifiers

This neuron detects the experimental-condition or variable-descriptor terms (e.g. “normal,” “pre-,” “post-,” “short-term,” “type-specific,” “pH,” etc.) commonly used in scientific paper titles to name factors being studied.

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

why

-1.70

-1.52

 what

-1.52

 when

-1.46

 They

-1.39

They

-1.37

-1.36

."

-1.33

 must

-1.33

DOCTYPE

-1.30

POSITIVE LOGITS

mitsub

1.36

jaguar

1.35

lxt

1.34

Én

1.34

to

1.32

不同

1.31

不同的

1.27

Sasuke

1.27

異なる

1.26

saya

1.25

Activations Density 0.415%