INDEX

Explanations

neurons describe research actions

The main thing this neuron does is detect gerund or present‐participle verbs ending in “-ing.”

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

 there

-2.13

 also

-2.08

на

-1.99

-1.92

and

-1.91

 seems

-1.77

''

-1.75

 sebagainya

-1.71

 found

-1.66

for

-1.62

POSITIVE LOGITS

the

2.97



1.92



1.90

 jednej

1.82

 them

1.80

 spezi

1.76

玙

1.73

粁

1.70

嵓

1.66

 underlines

1.64

Activations Density 0.127%