INDEX

Explanations

I, then, this, we, it

This neuron lights up on the first word of a new sentence (typically a capitalized pronoun, conjunction, or article at sentence start).

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

 burbu

-1.84

 magnifiques

-1.77

 incroy

-1.62

壠

-1.60

禩

-1.57

 contesta

-1.54

 betek

-1.53

 célèbres

-1.53

 bunda

-1.50

Maui

-1.50

POSITIVE LOGITS

 then

3.05

to

2.16

</strong>

1.98

now

1.85

 subsequently

1.73

 when

1.66

1.61

as

1.60

 also

1.57

1.52

Activations Density 0.022%