INDEX

Explanations

tensorflow, python, repos, fibrosis

The neuron activates on numeric tokens—especially floating‐point numbers.

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

 αυτό

0.91

 ช่วง

0.87

 único

0.82



0.82

 จุด

0.80

 solcher

0.80

 festgestellt

0.79

atributo

0.76

 weiterer

0.75

 nedeniyle

0.75

POSITIVE LOGITS

怙

0.84

examine

0.77

넬

0.77

ྦ

0.75

وى

0.74

جيل

0.73

نين

0.72

WHY

0.72

Existence

0.72

Activations Density 0.000%