INDEX

Explanations

depending on the

This neuron detects phrases expressing conditional variation, most prominently the word “depending” (as in “depending on the …”).

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

𓁹

-1.44

tituted

-1.38

𖣔

-1.34

should

-1.28

𐬹

-1.27



-1.26

taires

-1.25



-1.22

 должен

-1.21

 quintessential

-1.21

POSITIVE LOGITS

 their

1.48

of

1.41

it

1.40

you

1.31

 your

1.20

 each

1.19

 specific

1.13

 there

1.11

1.06

 circumstances

1.04

Activations Density 0.032%