INDEX

Explanations

numbers and relative positions

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ohn

-0.08

çī¹çĤ¹æĺ¯

-0.07

 '=',

-0.07

Ð°Ð¹

-0.06

ä»¥ä¸Ģç§į

-0.06

:]

-0.06

èĢħæĺ¯

-0.06

hood

-0.06

ahn

-0.06

èģĮä¸ļçļĦ

-0.06

POSITIVE LOGITS

hart

0.08

lexport

0.07

_-_

0.07

umul

0.07

ordion

0.07

embl

0.07

reator

0.07

à¹īà¸²à¸¡

0.07

rames

0.07

antung

0.07

Activations Density 0.002%