INDEX

Explanations

company names and banks

The neuron detects proper‐noun brand, company, and bank names (often in uppercase or acronym form).

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

ᥬ

-2.58

Ꮓ

-1.93

 parfüm

-1.92

禔

-1.91

colgante

-1.91

莯

-1.84

谂

-1.81



-1.80

掼

-1.80

 konten

-1.79

POSITIVE LOGITS

to

2.17

2.16

in

2.02

 have

1.93

 should

1.90

Of

1.88

 believes

1.86

 could

1.83

 They

1.83

 shows

1.79

Activations Density 0.004%