INDEX

Explanations

make : sentence

This neuron fires strongly on numeric tokens and list/section markers (e.g. digits, numbered steps, or double-hyphen separators).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 […]

0.77

",

0.69

<0xC2>

0.69

”,

0.68

]",

0.67

”

0.66

 {}",

0.65

[…]

0.65

0.63

',

0.61

POSITIVE LOGITS

↵↵↵↵↵↵

4.06

↵↵↵↵↵↵↵

4.03

↵↵↵↵↵

3.92

↵↵↵↵↵↵↵↵

3.91

↵↵↵↵↵↵↵↵↵

3.89

↵↵↵↵

3.81

↵↵↵↵↵↵↵↵↵↵

3.75

↵↵↵↵↵↵↵↵↵↵↵

3.73

↵↵↵↵↵↵↵↵↵↵↵↵↵

3.70

↵↵↵↵↵↵↵↵↵↵↵↵

3.69

Activations Density 0.207%