INDEX

Explanations

import statements source

The neuron fires on ES-style import (and export default) statements, in particular on the “from” and “default” tokens that introduce module imports/exports.

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

 suprême

-2.05

垍

-1.96

öglich

-1.94

 malade

-1.93

曁

-1.89

焢

-1.82

llten

-1.80

-1.77

-1.76

 作り方

-1.75

POSITIVE LOGITS

in

1.97

</h2>

1.90

 celebrated

1.88

Our

1.84

</td>

1.82

It

1.77

𝙖

1.77



1.76

 above

1.73

 fiabilidad

1.73

Activations Density 0.019%