INDEX

Explanations

Introduction or tag

The neuron consistently fires on tokens that end with the letters “al.”

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

-1.81

-1.66

To

-1.64

 acceptez

-1.61

-1.60

-1.58

of

-1.55

\\

-1.55

-1.54

-1.49

POSITIVE LOGITS

饃

2.00

>>>>>>

1.77

 eléct

1.71

>>>>>>>>

1.66

鲥

1.65

 現貨

1.62

 spesifik

1.57

 disiplin

1.57

Ꮉ

1.56

;;;;

1.55

Activations Density 0.000%