INDEX

Explanations

give you, do you, DOES THAT

The neuron strongly activates on content verbs—especially those expressing actions or judgments (e.g. take, give, think, undervalue, fools)—i.e. it’s looking for verbs that convey evaluation or directed action.

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



-1.35

RIQUE

-1.34



-1.34

銛

-1.31

 manuelle

-1.30

 superstars

-1.29

AWARDS

-1.28

杁

-1.28

 recommen

-1.27

ጢ

-1.27

POSITIVE LOGITS

you

1.82

him

1.51

 even

1.39

 others

1.39

ного

1.38

</h2>

1.33

 Vorgang

1.28

on

1.27

 Eindruck

1.26

ang

1.25

Activations Density 0.012%