INDEX

Explanations

Hindi, Russian, and Urdu words

This neuron responds to words or word fragments that are clearly non-English (foreign-language tokens or names).

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

𔘓

-2.13

 DEPOSITORY

-2.08

 jego

-1.67

 naszych

-1.66

 utilisée

-1.66

梣

-1.63

 его

-1.61

𝄆

-1.60

 producenta

-1.59

ίναι

-1.59

POSITIVE LOGITS

2.55

,’

2.20

’,

2.13

1.84

?’

1.77

’-

1.77

…’

1.77

.’

1.75

’:

1.75

’)

1.72

Activations Density 0.002%