INDEX

Explanations

date formats and numbers

The neuron primarily detects four‐digit year tokens (e.g. “2015,” “2023”) in the text.

New Auto-Interp

Configuration

Prompts (Dashboard)

238,145 prompts, 512 tokens each

Dataset (Dashboard)

lmsys + oasst1

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ју

1.41

]|[

1.40

(<

1.39

 FODC

1.38

aparikkh

1.38

$[\

1.38

 keras

1.34

bathtub

1.33

нің

1.32

 popupButton

1.31

POSITIVE LOGITS

vorschau

1.63

en

1.59

ي

1.55

væ

1.52

1.50

1.48

é

1.48

те

1.47

ak

1.42

स

1.37

Activations Density 0.051%