INDEX

Explanations

listening to

The neuron is tuned to detect the word “listening” in the text.

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

gebras

-2.63

and

-2.61

er

-2.59

’

-2.50

ТЫ

-2.39

poptotic

-2.31

et

-2.30

導致

-2.25

人的

-2.19

鸯

-2.17

POSITIVE LOGITS

</h3>

3.53

3.13

).

2.88

 intently

2.75

")

2.53

 That

2.47

нутрен

2.47

that

2.39

2.36

 rằng

2.33

Activations Density 0.010%