INDEX

Explanations

@theta or mathematical symbols

The neuron consistently activates on numeric tokens and arithmetic expressions (digits, decimals, and math‐related symbols).

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

what

-2.13

-2.09

To

-2.09

where

-1.90

that

-1.88

 strangle

-1.83

when

-1.83

at

-1.78

from

-1.74

這樣

-1.73

POSITIVE LOGITS



2.23

ㅁ

2.03

↵

2.03

Спољашње

1.89

 årene

1.88

bezpiec

1.86

 แท

1.83

 året

1.82

↵↵

1.80

歺

1.79

Activations Density 0.051%