INDEX

Explanations

code documentation and markup

The neuron is triggered by decorative or structural delimiter lines made up of long runs of non-alphanumeric characters (for example blocks of asterisks or other repeated punctuation).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

子は

-1.32

日から

-1.27

子が

-1.13

場は

-1.10

</tbody>

-1.09

は約

-1.08

},

-1.08

 پنجم

-1.05

はかなり

-1.05

子さん

-1.04

POSITIVE LOGITS

 paliwa

1.25

理由は

1.24

 calitate

1.22

なと

1.20

ところです

1.20

 singur

1.19

ところが

1.13

と思ったら

1.08

大きさ

1.05

考え方

1.05

Activations Density 0.010%