INDEX

Explanations

covariates, evaluator, evaluable

The neuron is strongly activated by stems of technical or methodological terms common in scientific writing (e.g. “gravit-”, “defen-”, “evalu-”, “covariates”), effectively picking out academic jargon.

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

 neues

-2.83

他们是

-2.61

戆

-2.59

我们是

-2.58

接下来的

-2.53

 company

-2.45

])){

-2.39

↵↵

-2.39

悫

-2.34

 męskie

-2.33

POSITIVE LOGITS

ان

2.98

 There

2.61

 وأ

2.33

This

2.28

up

2.27

to

2.25

 και

2.23

These

2.22

最

2.22

 pleas

2.17

Activations Density 0.006%