INDEX

Explanations

fruits and periods

The neuron strongly activates on tokens that appear inside parentheses—e.g. numeric values, format‐string keys, or titles bracketed by “(…)”.

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

,}$

0.76

,}

0.75

:',

0.73

...">

0.71

:}

0.71

?',

0.70

*}

0.69

!',

0.69

;'>

0.67

]-->

0.66

POSITIVE LOGITS

↵↵↵

1.06

⌋

1.01



1.00

〕

0.96

\)

0.96

</em>

0.93

right

0.87

﹚

0.86

``.

0.85

］

0.85

Activations Density 0.113%