INDEX

Explanations

compound nouns and phrases

The neuron selectively activates on numeric tokens (multi‐digit numbers or number‐like strings).

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

 FRUIT

-0.81

 almohada

-0.81

prc

-0.80

mw

-0.79

idc

-0.78

 ATLANTIC

-0.78

 tiroirs

-0.76

 habia

-0.76

städ

-0.76

 slaw

-0.74

POSITIVE LOGITS

 Without

1.17

 Anonymous

1.06

 Doctors

1.05

 Against

1.03

 without

0.92

 Gamb

0.88

Without

0.88

 Unlimited

0.88

 без

0.87

Anonymous

0.86

Activations Density 0.034%