INDEX

Explanations

enumeration markers

The neuron strongly activates on sentence boundaries—especially end‐of‐sentence punctuation (periods) and the first words of new sentences.

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

 will

-1.73

and

-1.73

 only

-1.54

 infine

-1.34

<eos>

-1.30

 cannot

-1.29

 Lastly

-1.27

最后

-1.22

 ultimately

-1.16

Finalmente

-1.16

POSITIVE LOGITS

Secondly

1.20

的同时

1.20

Specifically

1.20

 пър

1.16

 Secondly

1.15

Firstly

1.13

 באופן

1.13

 古い

1.12

 commencent

1.11

鴴

1.11

Activations Density 0.089%