INDEX

Explanations

helicopter and helicopters

The neuron strongly activates on mentions of helicopters (and related rotorcraft terms).

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

 considération

-2.58

他们的

-2.42

 leur

-2.42

 свою

-2.30

 offizielle

-2.23

℔

-2.14

.”

-2.11

 vostre

-2.08

𖤍

-2.06

ING

-2.05

POSITIVE LOGITS

3.16

);

2.97

It

2.89

 Bereichen

2.42

 разнообраз

2.36

}$

2.28

\\

2.23

驀

2.22

)-

2.19

atother

2.19

Activations Density 0.003%