INDEX

Explanations

was or had

The neuron detects tokens written in Cyrillic (i.e. Russian text).

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

or

-1.43

繕

-1.03

you

-0.98

或者

-0.98

 или

-0.97

unità

-0.96

 will

-0.94

넋

-0.94

üstü

-0.93

現在

-0.92

POSITIVE LOGITS

was

1.96

 hadn

1.69

 estava

1.47

had

1.41

 wasn

1.34

 était

1.21

でした

1.16

 Estaba

1.13

っていました

1.12

 jusque

1.10

Activations Density 0.054%