INDEX

Explanations

Override function calls

The neuron fires on the names of methods (especially override/declared function identifiers) in code.

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

THEY

-1.34

 intervi

-1.33

issent

-1.27

莇

-1.27

かなり

-1.27

𓐍

-1.24

 ruine

-1.22

 plafon

-1.20

𝗒

-1.20

 senin

-1.20

POSITIVE LOGITS

 that

2.25

 this

1.78

to

1.68

 there

1.59

 which

1.43

 while

1.35

良し

1.34

金

1.33

 most

1.33

 time

1.33

Activations Density 0.014%