INDEX

Explanations

used up or worn out

The neuron strongly activates on words describing mechanical wear and tear—especially variants of “worn/out” that signal something has become used up or unusable.

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

to

-2.52

</em>

-1.91

だが

-1.73

最为

-1.72

 limiti

-1.70

</td>

-1.66

 denies

-1.64

 είναι

-1.63

 remains

-1.59

踯

-1.59

POSITIVE LOGITS

</h1>

2.23

 Whether

1.74

 Moreover

1.71

 Aufenthalt

1.66

 Wettbewerb

1.65

 Where

1.63

॒

1.61

 Vorgang

1.59

 Spaß

1.58

 Glaube

1.58

Activations Density 0.005%