INDEX

Explanations

sequence followed by actions or context

The neuron fires on technical or domain-specific terminology—especially multi-syllable jargon, specialized units, and formal labels used in scientific, engineering, or professional contexts.

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

 that

-1.88

},

-1.36

 adecuadas

-1.34

 любой

-1.32

fatter

-1.30

 fazla

-1.30

 любы

-1.30

 любое

-1.29

 cálido

-1.29

 limpio

-1.28

POSITIVE LOGITS

for

1.65

masını

1.41

 железнодоро

1.32

to

1.30

 then

1.27

 absoluta

1.27

 henceforth

1.26

后来

1.24

achite

1.24

 kemudian

1.21

Activations Density 0.255%