INDEX

Explanations

hate, hatred, hateful things

The neuron detects mentions of hate or hateful language—words referring to “hate,” “hate speech,” “racist” messages, and similar expressions of bigotry.

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

५

-2.02

 включая

-2.02

ם

-1.91

’

-1.88

१

-1.88

zembro

-1.88

豋

-1.88

da

-1.84

عاية

-1.84

로

-1.84

POSITIVE LOGITS

of

2.30

2.17

鵟

2.14

↵

2.06

2.00

 stary

1.93

 stalwart

1.88

 tinham

1.86

 forlorn

1.81

 voient

1.80

Activations Density 0.004%