INDEX

Explanations

[reaction: see text]

The neuron strongly activates on the “[" character (left square bracket), i.e. it detects opening square brackets.

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

-1.77

的呢

-1.59

Despite

-1.56

After

-1.53

的吧

-1.48

 Despite

-1.47

 After

-1.45

so

-1.43

 Erwä

-1.42

 longstanding

-1.41

POSITIVE LOGITS

 interessante

1.62

 certaines

1.61

 visse

1.41

 enorme

1.38

が見える

1.34

witches

1.33

 GRAND

1.33

 insanely

1.33

 dando

1.32

interesting

1.30

Activations Density 0.005%