INDEX

Explanations

Base followed by specific types

This neuron activates on the token “Base,” identifying occurrences of “Base” in class or type names.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 about

-1.81

for

-1.73

 Архивировано

-1.72

 with

-1.66

劭

-1.63

in

-1.61

+',

-1.58

But

-1.55

"",

-1.55

nante

-1.53

POSITIVE LOGITS

2.17

毎月

1.96

1.94

ఙ

1.87

弊社

1.85

毎週

1.84

今年の

1.70

 wöch

1.66

Similarly

1.65

 proficient

1.65

Activations Density 0.010%