INDEX

Explanations

decorators and exports

The neuron activates on standalone numeric tokens (i.e. integer literals).

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

 even

-1.41

そして

-1.32

if

-1.23

meisters

-1.17

 ибо

-1.16

-1.14

Согласно

-1.13

</tr>

-1.13

 each

-1.10

-1.09

POSITIVE LOGITS

instead

1.23

感じで

1.20

 camiones

1.20

 cinturón

1.17

綺麗に

1.16

 détruit

1.15

 neón

1.13

様な

1.12

ところに

1.12

 veramente

1.11

Activations Density 0.005%