INDEX

Explanations

asking for information

The neuron spots interrogation cues—tokens like “if,” “why,” and “about”—marking when someone is asking a question or making a query.

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

 also

-1.91

mohon

-1.56

 always

-1.47

nė

-1.42

ÁT

-1.38

dė

-1.34

 sami

-1.34

to

-1.31

 with

-1.30

 kiek

-1.30

POSITIVE LOGITS

</h3>

1.98

句话

1.67

 about

1.64

葙

1.46

 чёр

1.45

頷

1.42

 ήταν

1.39

his

1.36

 evitando

1.35

 whoſe

1.33

Activations Density 0.014%