INDEX

Explanations

to find free idols

The neuron fires on verbs—especially auxiliaries and modals (e.g. must, may, will) and other action words indicating doing, obligation, or state.

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

Lá

-1.06

its

-1.03

our

-0.95

حف

-0.93

ongeza

-0.93

it

-0.89

 your

-0.88

腹部

-0.87

 вій

-0.87

)}$

-0.87

POSITIVE LOGITS

 himself

1.05

themselves

0.98

cended

0.96

 indicando

0.93

 seti

0.93

 vitamine

0.92

 DIFFERENT

0.91

 autonomie

0.91

essment

0.91

 garan

0.90

Activations Density 0.045%