INDEX

Explanations

The neuron is detecting numeric tokens (multi‐digit numbers and decimals) in the text.

Defendant, park, involvement

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

 konkret

-0.80

Gon

-0.79

 klimat

-0.78

Pla

-0.77

不该

-0.75

pla

-0.75

蹀

-0.75

見えます

-0.74

 Octa

-0.74

Pla

-0.73

POSITIVE LOGITS

via

0.99

 deadline

0.91

 downloading

0.90

 until

0.87

on

0.87

pośred

0.86

 áreas

0.85

 downloadable

0.85

槌

0.82

 razas

0.82

Activations Density 0.149%