INDEX

Explanations

committee

The neuron activates on occurrences of the word “committee,” flagging any mention of a committee.

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

 They

-2.17

 они

-2.11

ondheim

-2.02

鮑

-2.00

 roomy

-1.98

瑁

-1.98

 hotly

-1.97

 entire

-1.95

 большая

-1.93

价格

-1.92

POSITIVE LOGITS

肏

2.70

ですので

2.61

それにしても

2.59

the

2.56

というわけで

2.34

うーん

2.31

ܜ

2.22

2.19

虯

2.16

犰

2.14

Activations Density 0.006%