INDEX

Explanations

local data processing

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

é¹ħ

-0.05

ä¸Ģåİ»

-0.05

è¾¨è®¤

-0.05

apos

-0.05

ç¬¦åı·

-0.05

æµģ

-0.05

æŃ£

-0.05

ullets

-0.05

 wire

-0.04

.pos

-0.04

POSITIVE LOGITS

 localized

0.08

local

0.08

Local

0.08

 Local

0.07

 local

0.07

 granularity

0.07

-local

0.07

å±Ģéĥ¨çļĦ

0.07

 lokal

0.07

åĴĮåľ°æĸ¹

0.07

Activations Density 0.073%