INDEX

Explanations

famous people and athletes

The neuron fires on standalone capitalized tokens, i.e. proper nouns or acronyms.

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

he

-2.41

ist

-2.36

давно

-2.28

了一

-2.19

ſelves

-2.13

玉石

-2.11

。《

-2.08

 были

-2.06

ueroa

-2.05

 како

-2.02

POSITIVE LOGITS

筇

2.33



2.27

 obtenidos

2.09

 magnético

1.98

ゆく

1.97

锃

1.93

 trov

1.89

埽

1.88

และ

1.85

鰨

1.84

Activations Density 0.002%