INDEX

Explanations

so far

The neuron strongly activates on the discourse marker “so,” especially when it appears at the start of a clause.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

så

1.07

so

0.89

那么

0.88

那麼

0.85

så

0.80

Så

0.74

 וא

0.69

這麼

0.68

咁

0.66

 그렇

0.65

POSITIVE LOGITS

journ

1.30

othed

1.24

ledad

1.22

far

1.20

oooo

1.16

othes

1.13

jour

1.11

pped

1.10

UTHERN

1.02

pping

1.01

Activations Density 0.312%