INDEX

Explanations

place names and multi-lingual text

The neuron strongly fires on tokens that end in an “s” (including foreign‐language or possessive/plural/suffix forms).

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

of

-2.14

tonsoft

-1.96

 would

-1.91

ErrUnexpectedEOF

-1.79

モア

-1.77

ophageal

-1.76

StreetMap

-1.72

rimiento

-1.70

quatch

-1.64

 været

-1.59

POSITIVE LOGITS

own

2.53

 द्वारा

1.71

糴

1.70

 abilities

1.62

</h2>

1.61

és

1.60

 involvement

1.59

 και

1.56

んびり

1.56

 видит

1.55

Activations Density 0.007%