INDEX

Explanations

describing an idea or intention

This neuron detects the explanatory framing “the idea is…” (and similar “idea”-“is” constructions) that introduce a statement of purpose or intent.

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

螣

-1.62

齶

-1.38

and

-1.34

泥

-1.34

vertebrate

-1.31

布

-1.30

thentic

-1.29

_))

-1.29

trong

-1.27

hört

-1.27

POSITIVE LOGITS

 also

1.91

 only

1.69

 first

1.64

 eventually

1.56

想定

1.45

not

1.44

are

1.38

狙

1.38

get

1.34

if

1.34

Activations Density 0.053%