INDEX

Explanations

CJPR

This neuron fires on named imports (the identifiers listed inside `{ … }`) in JavaScript/TypeScript `import` statements.

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

 only

-1.23

Также

-1.16

owo

-1.15

all

-1.13

諍

-1.13

to

-1.10

 probably

-1.09

 sadar

-1.08

梟

-1.06

取締

-1.05

POSITIVE LOGITS

ふん

1.19

 reemplazar

1.15

↵↵

1.14

 pomocą

1.10

 mascotas

1.07

ようになる

1.05

 acesta

1.02

ええ

1.02

おわりに

1.02

 wcześniej

1.02

Activations Density 0.002%