INDEX

Explanations

median

The neuron fires on occurrences of the word “median.”

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

鲠

-2.80

</h3>

-2.78

Now

-2.70

 Jugendliche

-2.63

」、「

-2.59

家主

-2.56

Its

-2.56

nambuco

-2.55

 Therefore

-2.47

鹉

-2.34

POSITIVE LOGITS

皤

2.67

𖡼

2.41

翾

2.41

2.39

 invitamos

2.30

茕

2.30

来讲

2.23

うーん

2.22

涐

2.16

可谓

2.13

Activations Density 0.006%