INDEX

Explanations

programming and technical terms

This neuron does not respond to any of the tokens shown—it appears to be inactive (i.e. it does not detect any recurring pattern in these examples).

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

 Both

-2.33

or

-2.25

 During

-2.23

any

-2.11

and

-2.11

၆

-2.05

atação

-2.02

for

-2.02

 There

-1.92

婚礼

-1.91

POSITIVE LOGITS

 raiſ

1.80

說

1.70

1.66

耵

1.64

 apparaissent

1.64

 quintessential

1.64



1.63

 diminuer

1.63

1.60

镲

1.57

Activations Density 0.003%