INDEX

Explanations

pose challenges or describing states

This neuron is triggered by the character sequence “pose” (e.g. in words like “pose” or “poses”), regardless of context.

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

衕

-2.75

ization

-2.38

ༀ

-2.36

鉝

-2.31

縂

-2.31

趼

-2.23

ro

-2.22

ሷ

-2.20

 після

-2.17

雠

-2.14

POSITIVE LOGITS

𖣘

3.13

 certaines

2.45

années

2.31

 offrent

2.27

 voluntarios

2.23

</strong>

2.22

2.19

august

2.16

翆

2.14

涙

2.11

Activations Density 0.003%