INDEX

Explanations

originating from

The neuron strongly activates on the verb “originate” (and its inflected forms).

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

殽

-2.31

<bos>

-2.27

And

-2.22



-2.20

หรือ

-2.20

 señ

-2.14

￬

-2.09

where

-2.02

CURLOPT

-2.00

蠖

-2.00

POSITIVE LOGITS

潆

3.13

2.41

 overflowing

2.25

௜

2.17



2.14

 konnte

2.13



2.11

 kimdir

2.11

2.08

有不少

2.06

Activations Density 0.007%