INDEX

Explanations

code keywords and punctuation

The neuron fires on numeric tokens—especially floating‐point numbers and other multi‐digit numeric values.

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

 util

0.62

ourd

0.57

डक

0.55

 mains

0.54

 Алла

0.54

 Tatha

0.54

drm

0.54

ib

0.53

ապ

0.53

 ১৩

0.52

POSITIVE LOGITS

valueOf

1.04

ValueOf

0.64

愈

0.55

OMET

0.54

parseLong

0.54

parseDouble

0.54

TON

0.54

ENTO

0.52

Tera

0.52

गीय

0.52

Activations Density 0.012%