INDEX

Explanations

url and code snippets

The neuron flags rare, domain‐specific nouns or identifiers (e.g. library/client names, URL or code‐path segments, parameter or table names, scientific terms).

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

편

-0.72

 WORLD

-0.70

Perc

-0.68

 Avid

-0.68

anche

-0.66

REST

-0.66

ুন

-0.65

 aplicación

-0.65

WORLD

-0.65

чя

-0.65

POSITIVE LOGITS

来越

0.72

Tomo

0.63

راضي

0.63

cías

0.61

())));

0.61

 Prie

0.61

 Об

0.60

唷

0.60

 gyere

0.60

idx

0.60

Activations Density 0.080%