INDEX

Explanations

one such

The neuron detects mentions of a single, specific example—i.e. nouns introduced by words like “one” or “such.”

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

and

-2.09

 також

-1.23

もいい

-1.14

 också

-1.13

除了

-1.13

 壁纸

-1.12

🧍

-1.12



-1.09

≠

-1.08

 savait

-1.07

POSITIVE LOGITS

 dokonce

1.80

 даже

1.66

 even

1.64

 jopa

1.61

 sogar

1.60

 addirittura

1.56

 حتی

1.53

甚至

1.46

 nawet

1.34

 incluso

1.30

Activations Density 0.048%