INDEX

Explanations

difficulties or hardships

The neuron responds to words describing problem states or symptoms—i.e. terms that signal someone or something is unwell, stressed, failing, or otherwise in trouble.

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

看了一眼

-1.53

 poë

-1.52

 indispens

-1.43

来到了

-1.41

..........

-1.37

approximately

-1.35

quakes

-1.34

܊

-1.30

described

-1.29

摸了

-1.29

POSITIVE LOGITS

or

2.83

 или

1.75

 find

1.59

we

1.59

 focusing

1.58

 quickly

1.57

 helps

1.55

หรือ

1.54

 give

1.53

our

1.52

Activations Density 0.124%