INDEX

Explanations

hobbies and time-related events

The neuron is especially sensitive to words naming leisure activities or hobbies.

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

嫫

-2.03

好きです

-2.02

少ない

-1.98

 condenado

-1.97

 silencioso

-1.93

noche

-1.90

—

-1.86

 This

-1.86

𐑂

-1.83

 aided

-1.81

POSITIVE LOGITS

2.61

2.58

2.45

しました

2.33

 brooding

2.30

 нескольких

2.22

ll

2.19

྅

2.16

Этот

2.08

ные

2.05

Activations Density 0.003%